an alternative look at balance-of-payments puzzle: … · 2017-02-14 · 1 . robust and uniform...
TRANSCRIPT
AN ALTERNATIVE LOOK AT BALANCE-OF-PAYMENTS PUZZLE: STRUCTURAL DECOMPOSITION OF ACCOUNTS OF 16 EMERGING
MARKETS
A Master’s Thesis
by USSAL ŞAHBAZ
Department of Economics ......................................................
Bilkent University Ankara
January 2006
To the memory of Fatma Us (1924 – 2004)
AN ALTERNATIVE LOOK AT BALANCE-OF-PAYMENTS PUZZLE: STRUCTURAL DECOMPOSITION OF ACCOUNTS OF 16 EMERGING
MARKETS
The Institute of Economics and Social Sciences of
Bilkent University
by
AHMET USSAL ŞAHBAZ
In Partial Fulfilment of the Requirements for the Degree of MASTER OF ARTS
in
THE DEPARTMENT OF ECONOMICS BİLKENT UNIVERSITY
ANKARA
January 2006
I certify that I have read this thesis and have found that it is fully adequate, in scope and in quality, as a thesis for the degree of Master of Arts in Economics. --------------------------------- Prof. Erinç Yeldan Supervisor I certify that I have read this thesis and have found that it is fully adequate, in scope and in quality, as a thesis for the degree of Master of Arts in Economics. --------------------------------- Asst. Prof. Selin Sayek Böke Examining Committee Member I certify that I have read this thesis and have found that it is fully adequate, in scope and in quality, as a thesis for the degree of Master of Arts in Economics. --------------------------------- Assoc. Prof. Cem Somel Examining Committee Member Approval of the Institute of Economics and Social Sciences --------------------------------- Prof. Erdal Erel Director
iii
ABSTRACT
AN ALTERNATIVE LOOK AT BALANCE-OF-PAYMENTS PUZZLE: STRUCTURAL DECOMPOSITION OF ACCOUNTS OF 16 EMERGING
MARKETS
Şahbaz, Ahmet Ussal
Master of Economics
Supervisor: Prof. Erinç Yeldan
January 2006
The empricial evidence on pro-growth effect of capital account liberalization is
inconclusive. I argue that, after liberalization, the link between structural finance
needs of developing countries and debt flows lost its importance. Instead, the flight of
resident capital and unproductive reserve accumulation have created new financing
needs, limiting the augmentation of saving pools of developing countries and hence
growth. I build four new components from balance-of-payments account that make it
possible to track the new financing patterns. I investigate relationship between these
components for 15 emerging market countries. I also make a case study for Turkey
using vector autoregression technique to establish a causality link between dynamics
of the new structural aggregates.
Keywords: Balance of payments, emerging markets, financial crisis
iv
ÖZET
ÖDEMELER DENGESİ BİLMECESİNE YENİ BİR BAKIŞ:
16 YÜKSELEN PİYASA EKONOMİSİ HESAPLARININ YAPISAL
AYRIŞTIRILMASI
Şahbaz, Ahmet Ussal
Yüksek Lisans, İktisat Bölümü
Tez Yöneticisi: Prof. Dr. Erinç Yeldan
Ocak 2006
Sermaye hesabının serbestleştirilmesinin büyüme üzerindeki etkisine ilişkin ampirik
çalışmalar sağlam bir bulgu ortaya koyamamaktadır. Bu çalışmada,
serbestleştirmeden sonra gelişmekte olan ülkelerin yapısal finansman ihtiyacıyla bu
ülkelere olan sermaye akımları arasındaki ilişkinin önemini yitirdiği öne
sürülmektedir. Bu ilişkinin yerine, yerleşik sermayenin kaçışı ve rezerv birikimi
sonucu ortaya çıkan yeni finansman biçimi gelişmekte olan ülkelerin tasarruf
havuzlarının büyümesini ve dolayısıyla büyüme oranlarında artışı önlemiştir. Bu
yeni finansman biçimini gözlemlemek için ödemeler dengesi dört yeni bileşene
ayrılmakta ve bu bileşenlerin ilişkileri 15 yükselen piyasa özelinde incelenmektedir.
Ayrıca, Türkiye’ya dair bir vaka analizi yapılmakta ve vektör otoregresyon tekniği
kullanılarak bu bileşenler arasındaki nedensellik ilişkisi de araştırılmaktadır.
Anahtar Kelimeler: Ödemeler dengesi, yükselen piyasalar, finansal kriz
v
ACKNOWLEDGEMENTS
I thank Dr. Erinç Yeldan for his guidance through the development of this thesis. His
encouragement brought me to Bilkent for graduate study and his supervision made
the accomplishment of this thesis possible.
I am grateful to my examining committee members, Dr. Selin Sayek Böke and Dr.
Cem Somel. Their valuable suggestions have resulted in a great improvement in my
thesis.
I am indebted to many people who provided support in course of this study. Among
them, Dr. Korkut Boratav provided insight about the balance-of-payments system,
Fikret Karabudak and Hakan Toprak helped on data issues and Onur Mumcu
reviewed the last draft and corrected editorial mistakes. I am also grateful to Dr.
Kivilcim Metin-Ozcan, Dr. Umit Ozlale, Dr. Hasan Ersel, Dr. Bilin Neyaptı and
participants of Ecomod 2005 conference for their comments and suggestions.
I also thank Burcu Afyonoğlu, Barış Esmerok, Sırma Kollu, Bahar Tözün, Ozan
Acar and Tural Hüseyin of Bilkent Economics graduate class of 2005 for their
friendship and support during various stages of my graduate study.
Lastly, I thank my parents, Ferhunde Us and Abdurrahim Şahbaz, for encouraging
me to involve in academic studies and their support during all stages of my
education.
TABLE OF CONTENTS
ABSTRACT ..............................................................................................................III
ÖZET ........................................................................................................................ IV
ACKNOWLEDGEMENTS ........................................................................................V
TABLE OF CONTENTS ..........................................................................................VI
LIST OF TABLES .................................................................................................... IX
LIST OF FIGURES ..................................................................................................XI
1. INTRODUCTION................................................................................................... 1
2. MOTIVATION ....................................................................................................... 4
2.1. RECENT TRENDS IN CAPITAL FLOWS AND LIBERALIZATION................................... 4
2.2. EFFECTS OF CAPITAL ACCOUNT LIBERALIZATION ON GROWTH ........................... 14
2.2.1. The case for liberalization.................................................................... 14
2.2.2. Review of empirical evidence.............................................................. 17
2.2.3. The reasons of the failure of the case for liberalization ....................... 21
2.3. THE FLIP-SIDE OF BALANCE OF PAYMENTS........................................................... 25
2.3.1. Capital Outflows .................................................................................. 25
2.3.2. Reserve accumulation .......................................................................... 28
vi
3. METHODOLOGY................................................................................................ 32
3.1. BALANCE OF PAYMENTS ACCOUNTS AND THE TRADITIONAL REPRESENTATION....... 32
3.2. A NEW STRUCTURAL DECOMPOSITION ....................................................................... 39
3.3. DATA AND SAMPLE SELECTION ............................................................................ 46
4. LIBERALIZATION EXPERIENCES .................................................................. 51
4.1. SOME STYLIZED FACTS ............................................................................................... 51
4.2. COUNTRY ANALYSIS................................................................................................... 58
4.2.1. Argentina.............................................................................................. 59
4.2.2. Mexico ................................................................................................. 62
4.2.3. Chile ..................................................................................................... 65
4.2.4. Brazil .................................................................................................... 68
4.2.5. Peru ...................................................................................................... 72
4.2.6. Columbia .............................................................................................. 74
4.2.7. Venezuela............................................................................................. 75
4.2.8. Korea .................................................................................................... 77
4.2.9. Malaysia ............................................................................................... 80
4.2.10. Thailand ............................................................................................... 82
4.2.11. Indonesia .............................................................................................. 85
4.2.12. Philippines............................................................................................ 87
4.2.13. Hungary................................................................................................ 89
4.2.14. Czech Republic .................................................................................... 91
vii
4.2.15. Poland................................................................................................... 92
4.3. CONCLUDING REMARKS....................................................................................... 94
5. A CASE STUDY: TURKEY ................................................................................ 97
5.1. A STRUCTURAL ANALYSIS OF TURKISH BALANCE OF PAYMENTS ACCOUNTS ........ 97
5.2. TURKEY'S EXPERIENCE WITH AN OPEN CAPITAL ACCOUNT ................................... 101
5.3. CONCLUDING REMARKS........................................................................................... 109
6. CONCLUSION ................................................................................................... 111
SELECT BIBLIOGRAPHY .......................................................................................... 113
APPENDIX. CHRONOLOGY OF LIBERALIZATION EXPERIENCES ........... 119
viii
LIST OF TABLES
Table 1. List of Sample Countries............................................................................. 48
Table 2. External Debt Stock / GDP ratios ............................................................... 55
Table 3. Usage of Debt Flows in Sample Countries (1990 – 1996) ......................... 57
Table 4. Correlation of DF with other components in Argentina ............................. 62
Table 5. Usage of Debt Flows in Argentina (1990 – 1998) ...................................... 62
Table 6. Correlation of DF with other components in Mexico................................. 64
Table 7. Usage of Debt Flows in Mexico (1990 – 1994).......................................... 65
Table 8. Correlation of DF with other components in Chile..................................... 68
Table 9. Usage of Debt Flows in Chile (1991 – 1997) ............................................. 68
Table 10. Usage of Debt Flows in Brazil (1992 - 1998)........................................... 71
Table 11. Correlation of DF with other components in Brazil ................................. 71
Table 12. Usage of Debt Flows in Columbia (1993 - 1999)................................... 75
Table 13. Correlation of DF with other components in Korea ................................. 80
Table 14. Usage of Debt Flows in Korea (1990 – 1996) .......................................... 80
Table 15. Usage of Debt Flows in Thailand (1987 – 1996)...................................... 85
Table 16. Usage of Debt Flows in Indonesia (1990 – 1996) .................................... 85
Table 17. Correlation of DF with other components in Philippines ......................... 88
Table 18. Usage of Debt Flows in Philippines (1989 – 1996) .................................. 88
ix
Table 19. Usage of Debt Flows in Hungary (1992 - 1998)....................................... 90
Table 20. Usage of Debt Flows in Poland (1997 – 2003)......................................... 94
Table 21. Usage of debt flows in Turkey (1981 – 2000) .......................................... 98
x
LIST OF FIGURES
Figure 1. Capital flows to high-income OECD and emerging market countries ........ 5
Figure 2. Domestic and international financial liberalization................................... 10
Figure 3. Traditional Representation of Balance of Payments ................................. 34
Figure 4. The structural components of the Balance of Payments............................ 40
Figure 5. Debt flows and external debt stock for the sample countries .................... 47
Figure 6. Ratio of BOP components to external debt stock: All countries .............. 52
Figure 7. Debt flows to Asia, Latin America and East Europe (bil. $)..................... 54
Figure 8. Ratio of BOP components to external debt stock: Latin America............. 56
Figure 9. Ratio of BOP components to external debt stock: Asia ............................ 56
Figure 10. Ratio of BOP components to external debt stock: East Europe .............. 58
Figure 11. Ratio of BOP components to external debt stock: Argentina.................. 60
Figure 12. Ratio of BOP components to external debt stock: Mexico ..................... 63
Figure 13. Ratio of BOP components to external debt stock: Chile ......................... 66
Figure 14. Ratio of BOP components to external debt stock: Brazil ........................ 71
Figure 15. Ratio of BOP components to external debt stock: Peru .......................... 73
Figure 16. Ratio of BOP components to external debt stock: Columbia .................. 75
Figure 17. Ratio of BOP components to external debt stock: Venezuella................ 77
Figure 18. Ratio of BOP components to external debt stock: Korea ....................... 79
Figure 19. Ratio of BOP components to external debt stock: Malaysia ................... 81
xi
Figure 20. Ratio of BOP components to external debt stock: Thailand ................... 83
Figure 21. Ratio of BOP components to external debt stock: Indonesia .................. 85
Figure 22. Ratio of BOP components to external debt stock: Phillipines................. 87
Figure 23. Ratio of BOP components to external debt stock: Hungary.................... 90
Figure 24. Ratio of BOP components to external debt stock: Czech Republic ........ 91
Figure 25. Ratio of BOP components to external debt stock: Poland....................... 93
Figure 26. Ratio of BOP components to external debt stock: Turkey ...................... 98
Figure 27. Acc. impulse response functions (DF and KO, short lag length) .......... 103
Figure 28. Acc. impulse response functions (DF and KO, long lag length) ........... 105
Figure 29. Acc. impulse response functions (DF and BB) ..................................... 106
Figure 30. Rate of financial arbitrage in Turkey (%).............................................. 108
xii
CHAPTER I
INTRODUCTION
The surge in capital flows to the so-called “emerging market” countries1 has been
impressive in 1990s. A basic triggering element for this surge has been capital account
liberalization in those countries. It has been the argued that liberalization would be
beneficial for growth by attracting the idle savings in the industrialized countries to the
developing countries and thus narrowing the investment – saving gap in these countries
at a low cost of capital.
However, that has not been the case. Instead, capital account liberalization brought
volatility to growth and frequent crisis to emerging markets. Reviewing the extensive
cross-country literature on the issue, Prasad at al. (2003) stated that "there is no strong,
1 The countries which liberalized their capital accounts and attracted much capital inflows are called emerging markets. Although there does not exist a clear definition of an “emerging market” some reference lists may be informative: The Economist lists the following countries as “emerging markets”: China, Hong Kong, India, Indonesia, Malaysia, Philippines, Singapore, South Korea, Taiwan, Thailand, Argentina, Brazil, Chile, Colombia, Mexico, Peru, Venezuella, Egypt, South Africa, Israel, Czech Republic, Hungary, Poland, Russia and Turkey. Meanwhile, JP Morgan lists Brazil, Mexico, Russia, Turkey, Philippines, Venezuella, Colombia, Argentina, Peru, South Africa, Ecuador, Panama, Poland, Ukraine, Bulgaria, Nigeria, Egypt and Morocco in its “Emerging Market Bonds Plus – EMBI+ Index.” Morgan Stanley Capital International has a larger list: Argentina, Brazil, Chile, Chin, Colombia, Czech Rep., Egypt, Hungary, India, Indonesia, Israel, Jordan, Korea, Malaysia, Mexico, Morocco, Pakistan, Peru, Philippines, Poland, Russia, South Africa, Taiwan, Thailand, Turkey, Venezuela.
1
robust and uniform support for the theoretical argument that financial globalization per
se delivers a high rate of economic growth."
Attempts to identify the factors that link capital account liberalization to growth have
created a vast literature. This literature generally focuses on information asymmetries
and their negative results regarding attraction of more capital to emerging markets.
Nevertheless, flight of resident capital from emerging markets after liberalization, and
role of unproductive usage of capital flows as reserves in order to sustain capital flows
did not attract much attention. Liberalizing under macroeconomic imbalances,
inappropriate macroeconomic policies or poor management of capital account may
result in these unfavorable circumstances that may limit the pro-growth effects of capital
account liberalization even if the emerging market attracts much capital.
The structural financing needs of developing countries can be defined as current account
deficits plus net foreign direct investments. Before capital account liberalization, capital
inflows have traditionally been used to finance this structural need. I argue that after
capital account liberalization this relationship lost its importance. Instead, the flight of
resident capital and unproductive use of capital flows as reserves have created new
financing needs. Consequently, the role of capital account liberalization in narrowing
investment - saving gaps in emerging market countries have been limited.2
2 I elaborate other factors in limiting the positive effect of capital account liberalization on growth in Section 2.2.3.
2
The thesis is structured as follows: In the second chapter I describe the surge in capital
flows to emerging market countries in 1990s and its causes. I explain the case for capital
account liberalization. Then I review the empirical effects regarding the growth effects
of liberalization and I pinpoint some of the explanations for the failure of growth
premise. I also explain the role of capital flight and productive reserve accumulation in
this regard. In the third chapter, I describe my methodology. I give details about the
traditional presentation of balance of payments and the new components that will be
used in my analysis. The fourth chapter is the core of the thesis where I support my
argument by investigating the balance of payments dynamics of fifteen emerging market
countries. Lastly, in the fifth chapter, I provide a case study of Turkish financial
liberalization, using the same methodology in a historical framework of Turkish
economy in 1990s. I also provide vector autoregression results that establish a causality
relationship between capital inflows and outflows in this chapter. I conclude in the sixth
chapter.
3
CHAPTER II
MOTIVATION
This chapter provides a literature review and framework for the following analysis.
In the first section I overview the trends in capital flows in 1990s. In the second
section, I first outline the case for capital account liberalization. Then I review
empirical evidence regarding growth effects of it and lastly I outline some
explanations on the non-existence of a robust relationship between capital account
liberalization and growth. In the third section I turn the flip side of balance of
payments and suggest that capital outflows and reserve accumulation may also have
played a role in limiting the pro-growth effects of capital account liberalization.
2.1. Recent trends in capital flows and liberalization
The surge in capital flows to developing countries in early 1990s is a well-known
fact. The size of the surge is impressive. According to Eichengreen and Mussa
(1998) “net flows to developing countries have tripled from roughly $50 billion a
4
year in 1987 – 89 to more than $150 billion in each of the three most recent calendar
years.”
Figure 1. Capital flows to high-income OECD and emerging market countries
-50.000
0
50.000
100.000
150.000
200.000
250.000
300.000
350.000
400.000
450.000
1975 1980 1985 1990 1995 2000
High-income OECD countriesEmerging market countries
Source: IFS
High-income OECD countries: United Kingdom, United States, France, Ireland, Iceland, Netherlands, Japan, Norway, Switzerland, Sweden, New Zealand, Italy, Greece, Finland, Denmark, Germany, Canada, Belgium, Austria, Australia; Emerging market countries: China, Hong Kong, India, Indonesia, Malaysia, Philippines, Singapore, South Korea, Taiwan, Thailand, Argentina, Brazil, Chile, Colombia, Mexico, Peru, Venezuella, Egypt, South Africa, Israel, Czech Republic, Hungary, Poland, Russia and Turkey.
The figures are in million US dollars. Capital flows exclude foreign direct investment.
Figure 1 illustrates this magnitude of capital flows to developing countries in
comparison to high-income countries. Apparently, capital flows to high-income
5
countries also increased in 1990s, but this increase is surpassed by the surge in
capital flows to developing countries. For the countries represented in Figure 1,
capital inflows attracted by developing countries has increased 3.2 times from 1975
– 89 period to 1990 – 97 period, whereas capital inflows to high-income countries
was 10% higher in 1990 – 97 period compared to 1975 - 89.
The surge in capital flows today, albeit great, is not unprecedented in history. Global
capital markets witnessed another integration process approximately in the 40-year
period before 1914. The surge in the “first globalization wave” was larger than today
when the capital flows are scaled by GDP. However, there are some major
differences in composition of capital flows: The pre-1914 wave was dominated by
claims on governments, infrastructure and mining companies of developing
countries and those claims were mostly comprised of bonds. Conversely, today a
larger set of financial instruments are traded and claims on developing countries
involve equities as well as bonds (Bordo et al., 1998).
A similar comparison can also be undertaken with the expansion in capital flows in
1970s, largely thanks to the abandonment of Bretton Woods system. The 1970s rise
in capital flows was smaller compared to the recent surge, but it is worth reviewing
it to identify similarities and contrasts. The system through which capital flows was
directed at developing countries was operated in the following way: The trade
surpluses generated in oil-exporter countries were accumulated in banks of industrial
countries. This accumulation led to an expansion in private credit in the form of
6
syndicated bank loans to developing countries. Official development assistance,
which largely stemmed from cold-war politics, constituted the other way of capital
flows to developing countries in 1970s. Nevertheless, the debt crisis of early 1980s
in developing countries marked an end to this cycle. Commercial credit to
developing countries collapsed. Together with the effect of the fall in revenues of
oil-exporting countries in 1980s, this decade was characterized by a low level of
capital flows to developing countries (UNCTAD, 1999).
The surge in 1990s is not only larger than the size of flows in 1970s, but in many
aspects, it is structurally different. Firstly, private capital flows dominate official
flows. 80% of capital flows to developing countries are private in 1990s compared to
50% in the first half of 1970s. Secondly, within private inflows portfolio inflows
gained importance compared to the dominance of syndicated bank lending in 1970s.
As of 1997, the portfolio flows constituted more than a third of total flows to
developed countries whereas bank lending make up less than a third. In mid-1970s,
nearly three quarters of capital flows to developing countries were bank loans, the
remaining part being FDI. Thirdly, together with the switch to private flows from
official flows, the distribution of flows within developing countries becomes uneven.
Capital flows were concentrated towards some 20 countries, the so-called emerging
7
markets. The twenty countries3 which attracted 50% of capital flows in 1970 – 90
period have been the target of 90% of capital flows to developing countries in 1990s.
There are various factors that have contributed to this surge in capital flows in
1990s. Firstly, the change in demographic structure in industrial countries has
created a large savings pool to be directed at emerging markets. The accumulated
savings of baby-boomers has been a driving force of the surge in capital flows to
developing countries. Secondly, the development of institutional investors in
industrialized countries has provided a source for capital flows. Pension funds,
which accumulated the savings of baby-boomers, have invested to emerging markets
where return was high. Thirdly, financial liberalization in developed countries in
1980s made possible for these institutional investors to diversify their portfolio to
developing countries. With liberalization, the flexibility of banking system in
investment and credit creation increased, leading to financial innovation and
encouraging “securitization” – capital flows in the form of bonds and stocks.
Competition in financial services brought by liberalization also provided an
incentive to look for higher returns in probably more risky emerging markets.
Fourthly, technological innovation has facilitated this process. Advances in
information technology not only made it possible to access information on various
types of assets in minimum time, but also paved the way for creation of a broad set
of new financial assets such as options, futures, and swaps. Progress in information
3 Argentina, Brazil, Chile, China, Columbia, Ecuador, Egypt, India, Indonesia, Malaysia, Mexico, Morrocco, Peru, Philippines, South Korea, Thailand, Tunisia, Turkey, Uruguay, Venezuella.
8
technologies made it possible to trade assets all over the world simultaneously and
reduced information asymmetries to some extent by reducing the cost of getting
knowledge about far-distant countries. Lastly, and probably most importantly,
capital account liberalization in emerging markets has made possible for various
types of flows to get into these countries (UNCTAD, 1999; Eichengreen and Mussa,
1998)
Capital account liberalization has been an obvious trend in emerging market
countries since late 1980s. Capital account liberalization or international financial
liberalization4 has generally been part of a policy-package, which also includes
domestic financial liberalization and has been implemented after or together with
current account (or trade) liberalization. This policy package has sometimes
included other measures for domestic economic liberalization such as privatization
and deregulation.
In the literature analyzing financial liberalization, international financial
liberalization has only been one dimension, the others being other reform packages
targeted at domestic financial sector. Williamson and Mahar (1998) list elements of
financial liberalization as follows: (1) elimination of credit controls, (2) deregulation
of interest rates, (3) free entry into banking sector, or more generally, the financial
services industry, (4) bank autonomy and (5) private ownership of banks, and lastly
(6) international financial liberalization.
4 I use the terms “capital account liberalization” and “international financial liberalization” interchangeably throughout the thesis.
9
Figure 2. Domestic and international financial liberalization
0
0,5
1
1,5
2
2,5
1973 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 19950
2
4
6
8
10
12International (left axis)Domestic (right axis)
Source: Mody and Abiad (2005)
The figure illustrates the Average values for Mody / Abiad indexes for domestic and international reform categories. Mody and Abiad (2005) measures six dimensions of financial liberalization: (1) elimination of credit controls, (2) deregulation of interest rates, (3) free entry into banking sector, or more generally, the financial services industry, (4) bank autonomy and (5) private ownership of banks, and lastly (6) international financial liberalization. In their index 3 represents full liberalization and 0 represents full repression.. Domestic financial liberalization index is the total of the first 5 components, hence ranges from 0 to 15. International financial liberalization the sixth component and ranges from 0 to 3.
In many instances, reform in these six dimensions have been implemented together
or with short intervals. Mody and Abiad (2005) provide indexes for these
liberalization dimensions for 35 industrial and developing countries for the period of
1975 – 1996; where, in each dimension of reform “0” represents full repression and
“3” represents full liberalization (henceforth Mody / Abiad index). A graphical
illustration of averages of Mody / Abiad indexes of domestic and international
10
financial liberalizations in Figure 2 suggest that in general international and
domestic financial liberalization go hand-in-hand5. Kaminsky and Schmukler (2002)
also conform to this view.
In order to make the analysis clear, throughout the thesis, I differentiate international
financial liberalization from domestic financial liberalization and I use capital
account liberalization (or international financial liberalization / integration) as a set
of policies which include6: (1) measures that allow foreign residents to hold
domestic financial assets, either in the form of debt equity, (2) measures that allow
domestic residents to hold foreign financial assets, (3) measures that allow foreign
assets to be freely held and traded within domestic economy. It should be noted that
these classification measures, from (1) to (3), represent an increasing degree of
international financial liberalization; and there exist many countries which engaged
in capital account liberalization but has not implemented some measures within
these categories, especially in category 3.
It is hard to quantify international financial liberalization. Nevertheless, attempts to
this end in the literature show a general liberalization trend in emerging market
countries. In the most recent attempt, the average for Mody / Abiad index
5 Mody and Abiad support the view that domestic / international financial sector liberalization is stimulated by liberalization in other economic dimensions. They also point out the importance of regional diffusion, shocks related to new governments, shocks that emanate from balance of payments crisis are other factors that increase likelihood of financial liberalization. They also argue that trade openness is a structural factor associated with more financial liberalization. 6 This classification of reforms are due to Ghosh (2005).
11
(mentioned in the preceding paragraph) for the emerging market countries it covers
rises from 0.62 from 1980 to 0.96 in 1990 and 1.85 in 1996.
Williamson and Mahar (1998) make an extensive review of liberalization episodes
in various countries. They classify capital account regimes into four categories;
namely repressed, partially repressed, liberal and largely liberal. According to their
classification, of the 25 developing countries in their sample, from 1973 to 1996, 13
have stepped two or more steps up in liberalization, 5 stepped one step up in
liberalization and 7 already had liberal capital account regimes in 1973.
Quinn and Inclan (1997) build an index from legal restrictions presented in IMF
Annual Report on Exchange Arrangements and Exchange Restrictions covering
restrictions on capital account, export proceeds and multiple exchange rates, in
which “0” represents a closed economy while “12” represents an open economy.
Their index, as cited in Eichengreen and Mussa (1998), averages constantly around
2.5 for emerging markets between 1960 – 80, then average rises to 3 around 1985
and over 3.5 in 1990s.
Kaminsky and Schmukler (2002) construct another index by reviewing liberalization
chronology in 28 countries; and the average capital account liberalization index for
the 14 emerging markets in their sample falls to 1.5 at the end of 1990s from a
plateau over 2.5 in 1980s, where “3” represents full repression and “0” represents
full liberalization.
12
It should be noted that in these quantifications of capital account liberalization,
transition countries from communist regimes are omitted because no data is
available for these countries when their capital accounts were repressed. When
liberalization in transition countries is also taken into account, the quantitative
liberalization trends will be strengthened.
Meanwhile, controls on capital flows have also been a hot issue in the literature.
Some countries preferred implementing measures of control instead of full
liberalization. The most important example is the unremunerated reserve
requirements or encaje in Spanish, applied in Chile from 1991 to 1998. Encaje was
basically a control mechanism on inflows. As it necessitated a certain portion of
inflows to be deposited in non-interest bearing accounts at the Bank of Chile, it
played the role of an implicit tax on inflows. A similar control was also applied in
Columbia between 1993 – 97. Another sort of capital controls were implemented in
Malaysia and Thailand in 1997 after Asian crisis. This time, the controls were
directed on outflows of capital rather than inflows, and required portfolio outflows
to be held at non-interest bearing accounts for a certain period. Nevertheless, these
controls were measures against crisis and reversed in a period of one year. Literature
on capital controls suggest that, they have limited effect in constraining size of the
capital flows and gaining monetary policy independence to the countries where
controls are applied. They have been effective, however, in lengthening the maturity
of capital flows (Edwards, 1999 and de Gregorio, 2000).
13
The above review shows that capital account liberalization in developing countries
has been a trend since late 1980s, and has been one of the stimulating factors for the
other trend of the same era, the expansion of capital flows directed at emerging
markets.
2.2. Effects of capital account liberalization on growth
As it was one of the major driving forces behind the surge in capital flows to
developing countries, it is worth examining the motivation of those countries for
international financial liberalization. In this section, I first outline the case for
liberalization. Then, I review the empirical studies that investigate the effect of
capital account liberalization on growth and conclude that there exists no ample
evidence of a positive effect for emerging market countries. Lastly, I review some of
the reasons offered in the literature for the lack of such a positive relationship.
2.2.1. The case for liberalization
The basic argument of the theoretical case for capital account liberalization is the
augmentation of national savings via a more efficient allocation of world savings.
The argument goes as follows: The developing countries are capital-scarce; hence
they offer higher returns to savings. In the meantime, the rich countries are capital-
14
abundant; therefore they have lower returns on capital. With capital account
liberalization, the savings of rich countries will flow to developing countries where
they find high returns, and help them closing the investment – saving gaps (Laurent
et al., 2002). This process enables investors to achieve higher risk-adjusted returns
and income levels in recipient countries rise as a result of capital inflows. As
Eichengreen and Mussa (1998) state, “higher rates of return can encourage saving
and investment that deliver faster rates of economic growth.”7
Another argument supporting liberalization is the risk-sharing premise, which
basically states that capital account liberalization enables domestic agents in
developing countries to diversify their risks and hence reduce unfavorable domestic
cyclical effects. This better allocation of risk is expected to spur domestic
investment and lead to high growth rates. Domestic agents can also borrow from
abroad when incomes are low and lend when incomes are high and smooth their
consumption.
7 They draw on from Goldsmith (1969), McKinnon (1973) and Shaw (1973) when putting forward this argument. See also Lucas (1990) on the reasons deterring flow of capital from rich to poor countries. He suggests that the differences in returns to capital between rich and poor countries may disappear once the differences in human capital, external benefits of human capital, monopoly power on capital flows to certain jurisdictions and political risk associated to countries are taken into account.
15
It is also expected that with capital account liberalization, domestic financial sector
will develop in emerging market countries. A deeper the financial sector will be
another channel for international financial liberalization to enhance growth8.
Lastly, more prudent domestic macroeconomic policies are expected to be
implemented after capital account liberalization in order to sustain capital inflows.
This policy discipline effect is also predicted to increase domestic investment and
productivity, paving the way for higher growth rates (Eichengreen and Mussa,
1998).
Apparently, the case for capital account liberalization is very similar to the case for
current account liberalization. Free movement of capital is efficiency-enhancing like
free movement of goods, and hence promotes growth9. Although capital account
followed current account liberalization in many countries, or happened hand-in-hand
in some instances, the empirical results regarding the effect of the former on growth
is different from the latter. Most studies report a positive and robust relationship
between current account liberalization and growth, whereas the relationship of
capital account liberalization with growth is at least mixed. I review this literature in
the next section.
8 See King and Levine (1993) for the positive relationship between financial development and economic growth. 9 See Rodrik and Rodríguez (2001) for a skeptic view on association of growth and trade liberalization.
16
2.2.2. Review of empirical evidence10
In general, the significance of the effect of liberalization on growth is tested by
adding liberalization as an independent variable to an equation where growth is the
dependent variable and the other independent variables are standard variables in
growth literature, such as initial per capita income to measure convergence,
investment, population, secondary school enrollment rates and regional dummies.
Rodrik (1998) and Quinn (1997) are two pioneers of this literature who come up
with differing results concerning the effect of international financial liberalization on
growth. Rodrik uses a sample of one hundred industrial and developing countries for
the period of 1975 – 89, he finds no explanative power of capital account
liberalization on growth. Quinn, conversely, suggests that capital account
liberalization is positively associated with growth in the period of 1960 – 89 in the
64 developed / developing countries in his sample.
There are various differences between methodologies of Rodrik and Quinn which
also persist in the literature after them. Rodrik uses a de jure measure of
liberalization compiled from IMF Annual Report on Exchange Arrangements and
Exchange Restrictions. His measure of liberalization is binary, 0 meaning closed and
1 meaning open. On the other hand, Quinn builds a measure based on anecdotal
evidence provided in IMF Annual Report on Exchange Arrangements which ranges
10 I follow Eichengreen (2003) which makes an extensive review of cross-country studies investigating effects of capital account liberalization on growth.
17
from 0 – 4 with increments of 0.5. Many authors refer to Quinn’s measure as “more
informative.” Nevertheless, the difference in methodology is not limited to
liberalization measures. Quinn’s sample is more favorable to provide positive
evidence because it contains more industrial countries than Rodrik’s sample; and it
has a larger time span, lowering the weight of 1980s which developing countries
struggled with crisis and low growth. Lastly, Rodrik uses levels of capital account
liberalization, whereas Quinn uses changes in capital account liberalization. The
extensive literature after Rodrik and Quinn, which has been reviewed by Edison et
al. (2004), provide differing results concerning the effect of liberalization on growth.
The differences in outcomes are due to varieties in the sampling of countries, time
periods and forms of capital flows (FDI / portfolio / debt; public / private, etc.), and
use of levels of differences when considering capital account liberalization, and
other explanatory variables in the regression11, 12.
The idea that capital account liberalization may be conducive to growth in some
countries (developed ones) and not in the other countries (developing ones) has
created another segment of literature pioneered by Edwards (2001). Edwards uses
Quinn’s measure of liberalization (his results are also robust when Rodrik’s measure
11 Bekaert et al. (2001) focuses on growth effects of equity market liberalizations and report a robust positive relationship between growth and liberalization. It should be noted that this study does not cover liberalization in the portfolio investments on bonds and other investment aspects of capital account. 12 A non-econometric study by Gourinchas and Jeanne (2004), using a neoclassical growth model, predicts that switching from financial autarky to international integration is roughly equivalent to a one percent permanent increase in consumption, which is negligable compared to productivity gains that will stem from improving internal allocative efficiency.
18
is used) and reports that while capital account liberalization is positively associated
with growth in high-income OECD countries, the relationship vanishes when a
sample composed of emerging market countries is used. Edwards suggests that the
non-linearity may be a result of financial development in developed countries. Yet,
Arteta et al. (2001) criticize Edwards (2001) on methodological grounds; and instead
of a threshold of financial development, they find out that the positive effect of
liberalization may be associated with other measures of openness. Consequently,
they argue that sequencing is important to reap positive effects of liberalization and
current account liberalization should be accomplished before capital account
liberalization. Meanwhile, Prasad et al. (2003) also report some evidence of a
“threshold” effect, implying that the absorptive capacity of markets may have a role
in reaping benefits of international financial integration.
More recently, institutional quality has also been investigated as the cause of non-
linearity in the effect of capital account liberalization on growth13. Edison et al.
(2004) extend the view of Rodrik (1998) that capital account liberalization may
proxy for government reputation in a growth regression if the latter does not appear
as an explanatory variable. Using both Rodrik’s and Quinn’s measures (and some
other measures) they find that the positive relationship between capital account
liberalization and growth vanishes when government reputation is added as a control
variable.
13 Hall and Jones (1999) and Acemoglu et al. (2001) are pioneers in the literature that relates growth to institutions.
19
Lastly, another paper by Quinn et al. (2001) investigates effects of three variables
that may result in non-linearity: (1) economic states before liberalization in line with
Edwards (2001) and Arteta et al. (2001); (2) political and legal states before
liberalization with the suspicion that democratic regimes may have high
compensation costs for the losers from liberalization that may overweigh positive
effects; and (3) level of social development before liberalization. They find evidence
that liberalization is positively associated with growth but not for emerging market
countries with democratic regimes14,15 .
To sum up, cross-country literature reviewed here provides no ample evidence of
significant positive effect of capital account liberalization on growth. It is suggested
that the positive effect, if any, is limited to developed countries rather than
developing countries. Yet, the attempts to identify the causes of the threshold effect
have failed to provide decisive answers up to now. In the next section, I review some
of the reasons that are proposed to explain the absence of the link between capital
account liberalization and growth in emerging markets.
14 A recent paper by Alfaro et al. (2005) relates capital flows to the classical variables / instruments in institutions literature. They find capital flows are positively associated with institutional quality. Historical determinants of current legal institutions have a direct effect on capital flows. They also suggest that there is room for the role of policy in determining the level and volatility of capital flows. 15 Meanwhile, Edison et al. (2002) relate various measures of international financial integration to growth. These measures include Rodrik and Quinn’s de jure capital account liberalization measures and measures of capital inflows and inflows / outflows as a proxy for actual integration. They are unable to provide a robust link between growth and international financial integration even when they control financial, institutional and policy characteristics.
20
2.2.3. The reasons of the failure of the case for liberalization
The literature suggests various reasons why capital account liberalization did not
deliver its premise of growth. The major explanation lies in the difference between
the mechanisms through which capital markets and goods markets work. Unlike the
markets for goods, financial markets are typically characterized by “information
asymmetries.”16 The first outcome of this characteristic is the “adverse selection”
problem which occurs when lenders have incomplete knowledge of borrower
quality. Under adverse selection paradigm, lenders will lend at higher rates than the
rates deserved by “good” borrowers, thus good borrowers will have an incentive not
to seek loans, leaving the lenders with “bad” borrowers. The allocation of capital in
this framework, obviously, will not be efficient.
The second outcome of information asymmetries is known as “moral hazard”: The
borrowers have a chance to alter their behavior after they borrow. If the project goes
well, then the borrower will reap the benefits; but if the project goes bankrupt the
lender will bear the costs. As a result, the borrower will engage in excessively risky
behavior. This will cause the lenders to seek guarantees for their loans, and in many
instances the outcome will be explicit or implicit government guarantees on private
debt of emerging market countries. Consequently, with insurance on debt, the
outcome will be excessive lending for risky projects. It is worth noting that, many
emerging market financial systems are characterized by a mismatch between short-
16 The further explanations of information asymmetries draw on from Eichengreen and Mussa (1998) and Rodrik (1998).
21
term liabilities and long-term assets as a reflection of intrinsic information
asymmetries.
Lastly, information asymmetries may lead to “herding” behavior, where agents
follow the lead of those whom they believe to be better informed. As international
fund managers place too little weight to their private information, the likelihood of
herd behavior increases. The mismatches that stem from moral hazard increase
vulnerability of the system to runs17, rendering herding behavior of international
investors “rational.” Herding amplifies the effects of both favorable and unfavorable
shocks. The result of the former is excessive lending, whereas the result of the latter
may be a crisis. Moreover, as many international fund managers treat emerging
markets as a basket, rather than giving an individual treatment to each market,
herding behavior raises contagion effect of a crisis in one market to another. Short-
term nature of borrowing in emerging markets combined with herding behavior may
trigger “sudden stops” of capital inflows despite good macroeconomic
fundamentals18.
It is also worth noting that the information asymmetries that are endemic to financial
markets are amplified when financial transactions take place in international domain.
Obvious reason for this phenomenon is the increasing cost of gathering and
evaluating information with increasing geographic distance and cross-cultural
17 Rodrik and Velasco (1999) show that distorted incentives typically cause excessive short-term foreign borrowing, which causes runs at the end. The larger the short-term borrowing, the larger its real consequences regarding reduced output and consumption. 18 See Calvo (1998) for a detailed investigation of sudden stops.
22
differences. Moreover, enforcement of contracts is substantially harder when the
lender and borrower are in different jurisdictions compared to financial transactions
that take place in the same country.
The information asymmetries may hinder the pro-growth effect of capital account
liberalization through several channels. Firstly, information asymmetries result in an
inefficient allocation of capital. Consequently, market mechanism fails to provide an
efficient allocation of capital to the developing countries.
Secondly, as a result of information asymmetries the capital flows become pro-
cyclical, i.e., the funds are not available when there are weaknesses / there are more
funds than you wish to borrow when economy is going well. The pro-cyclical nature
of capital flows makes the risk-sharing and consumption smoothing channels
ineffective. Indeed, this channel works in the opposite way than expected.
Thirdly, the positive effect of financial deepening channel on growth becomes
flawed given volatile characteristics of capital flows. Financial deepening does not
per se imply efficient allocation of resources and growth. Instead, given the
information asymmetries elaborated above, financial deepening may result in
inefficiencies. Typically, domestic investors will not finance long-term investments
with the short-term borrowings available from international markets. As a result, the
financial expansion is more likely to finance government deficits or consumption
booms. In addition, typically most financial crises are preceded by a phase of
financial deepening.
23
In a similar way, the effects of capital account liberalization may be limited or even
welfare reducing in the presence of domestic distortions. A classical illustrative case
is capital account liberalization before current account liberalization: If a labor-
abundant country protects its capital-intensive industries, after international financial
liberalization, the funds will flow to this protected sector, in which the country has
no comparative advantage. Consequently, the misallocation of resources will be
intensified after liberalization. This allocative efficiency effect is not limited to trade
barriers but also will be present when there are other distortions such as
macroeconomic imbalances.
Lastly, policy discipline may not be imported with capital accounts liberalization.
With a typical surge in capital inflows after liberalization due to higher returns of the
emerging market country, the governments are more likely to engage in
irresponsible behavior, such as running larger fiscal deficits. The experience shows
that no one cares about imbalances as long as they can be sustained through capital
inflows, and this gives government to exacerbate imbalance. When a sudden stop of
inflows occurs, the imbalances result in a crisis.
Moreover, in most countries that liberalize capital accounts, as a consequence of a
surge in capital inflows, real exchange rate appreciates seriously, causing an
incentive to invest in non-tradables, and discouraging investments in tradables. It is
hard to manage the real appreciation effect and distortions caused by real
appreciations can lead to deindustrialization in developing countries (Ghosh, 2005).
24
The reasons that we examined in this section, which also have been the major
themes of the literature that explains the lack of association between capital account
liberalization and growth, miss an important point: To what extent capital imported
after liberalization has been offset by capital exported; and to which extent did the
intrinsic risks of liberalization pave the way for unproductive use of capital
imported. In the next section, I turn the flip side of the balance of payments accounts
to shed light on some possible dynamics that further breaks the link between
liberalization and growth.
2.3. The flip-side of balance of payments
In this section I will explain the outflows of resident capital and reserve
accumulation in developing countries after they liberalized their capital accounts. I
will outline the reasons, dimensions and costs of these two phenomena for emerging
economies. This section provides a brief explanation for the reasons that capital
outflows and reserve accumulation limited the growth promoting role of
international financial liberalization.
2.3.1. Capital Outflows
Most studies reviewed in Section 2.2 concentrate on capital inflows. Failure to
obtain any significant relationship between net capital inflows and growth in cross-
25
country regressions demonstrates the extent to which the financial market
imperfections played a role in limiting the channels put forward by case for capital
account liberalization. Nevertheless, when assessing the effect of capital account
liberalization to close the investment – saving gap in a developing country, one
should also consider outflow of resident capital. Put it in another way, a country that
attracts inflows of non-resident capital as much as it encourages outflows of resident
capital after the capital account has been liberalized will fail to augment its saving
pool. One can argue, by focusing on capital inflows, that this country has not
attracted capital after liberalization. However, in fact, the country has attracted non-
resident capital, meanwhile it has also lost its resident capital and the balance gives
zero. The policy implications for the two analyses are quite different: Focusing on
inflows, one can give advice to implement policies to attract more capital inflows
(perhaps by raising the interest rates). If one considers the inflows / outflow
perspective, the advice will be taking measures to decrease outflows in order to reap
benefits of inflows in closing investment – saving gap.
In an emerging economy, the risks faced by resident and non-resident investors are
typically different. Resident investors face inflation and exchange rate risks; whereas
the major risk that non-resident investors confront is the default risk. Both investors
face taxes, which include not only explicit taxes on capital, but also include inflation
(for residents) and default (for non-residents) taxes. For a government of a
developing country with a fiscal deficit, an easy revenue generation source can be an
unexpected increase in inflation. In this way, resident investors are taxed. If foreign-
26
currency-denominated assets are not available domestically, given chaotic monetary
history in many emerging markets, it is very likely for residents to acquire claims on
non-residents (Dooley, 1988). Moreover, lack of confidence in the enforcement of
the residents’ property rights may encourage them to diversify their portfolio to
foreign assets.
Once the restrictions on outflows are eased, given today’s high level of information
technologies and financial sophistication, it has been easy for emerging country
residents to diversify their portfolios to include foreign assets. According to
Cornford and Brandon (1999), “around 300 banking entities from 10 leading
developing countries were operating in OECD countries in 199619.
Arguably, capital outflows are beneficial for the home country, since the gross
national product (GNP) is maximized as a result of portfolio diversification
wherever the profits are earned. However, as Stiglitz (2000) argues, when there are
positive externalities from domestic investments, such as taxes on capital (which are
hard to apply to investments abroad), returns to scale or other spillovers, the utility
of domestic investments is higher than investments abroad. In this case, the objective
may be maximizing gross domestic product (GDP), not GNP. This case is especially
applicable in early stages of development.
19 cited in UNCTAD (1999: 107). The ideas on capital outflows are gathered from UNCTAD (1999) and Laurent et al. (2002).
27
2.3.2. Reserve accumulation
Investigations of net inflows or gross inflows have also limited power in
demonstrating the effect of international financial integration balancing investment –
saving deficit, when the end use of inflows is disregarded. As capital flows are
characterized by sudden stops and reversals, emerging markets need to establish
credibility to sustain them. The most important way to enhance credibility is to
accumulate reserves. Official reserves are seen as insurance to pay the short-term
debt of the country in case of a reversal in capital inflows; and ratio of reserves to
foreign currency denominated short-term debt has been established as firm
indicators of crisis in the literature (Rodrik and Velasco, 1999). When this ratio falls
below one for an emerging market country, it is likely that the worried international
lenders will exhibit a herding behavior in exiting from the country, culminating in a
crisis. Therefore, emerging market countries has to use a significant amount of
capital inflows to accumulate reserves.
This reserve accumulation motive to defend exchange rate and establish credibility
is unique to the period after financial liberalization. In a closed capital account
regime, as the capital flows are related to imports and current account financing, the
need to maintain reserves only arises from the time lags between payments of
imports and receipts of exports. According to UNCTAD (1999: 23), “traditionally,
reserves covering on average three or four months’ imports are considered as
adequate for such purposes, and even smaller reserves would be needed to the extent
28
that governments are more willing to respond to current account disturbances by
exchange rate adjustments.” As a result of international financial liberalization, the
number n in the statement “reserves cover n months of imports” has increased from
3.5 in 1980 to 5.5 in 1997 for developing countries.
The new standard of reserve level suggested by Alan Greenspan20 is accordingly
much higher than to cover several months of imports. As cited in UNCTAD (1999:
111), he tells “countries could be expected to hold sufficient liquid reserves to
ensure that they could avoid new borrowing for one year with a certain ex ante
probability, such as 95% of the time.” Feldstein (1999) also conforms with this view.
He put forward that “the most direct way for a country to achieve liquidity is to
accumulate substantial amounts of liquid foreign reserves.” He also adds “China’s
$140 billion in reserves sends a strong signal to investors”21.
However, reserve accumulation has opportunity costs for the emerging market
country. Reserves tie up purchasing power that could be used to import goods
needed for investment and increase output. Therefore, reserve accumulation has a
substantial effect in limiting the capital account liberalization’s effect on closing
20 Chairman of Federal Reserve Board. The citation is from his speech at World Bank conference on Recent Trends in Reserves Management, Washington, DC. 29 April 1999. 21 A function of reserve accumulation different from the precautionary motive against financial crisis is to prevent appreciation of local currencies and thus promote export competitiveness. China’s reserve accumulation, for example, is mainly explained by this motive. Nevertheless, when the general application in emerging economies is considered, the motive of accumulating reserves has generally been precautionary and maintaining competitiveness motive has not been important. See Aizenman and Woo (2005) for more detail.
29
investment – saving gap. Stiglitz (2000: 1081) explains the cost of holding reserves
in a simple manner:
...consider a poor developing country. A company within the country borrows,
say, $100 million from a US bank that charges him 20%. If the country has
been maintaining what it views as minimum prudential reserves then it will
have to add $100 million to reserves. For simplicity, assume it holds those
reserves in US T-bills. Consider the implications from the perspective of the
country’s balance sheet and income flows: It has lent the United States $100
million and borrowed from the United States the same amount – it has no new
capital. But it pays to the United States every year the $20 million in interest,
while it receives from the US $5 million, the interest on the T-bill. Clearly,
this is a good deal for the United States but it is hardly the basis for more rapid
growth by the poor developing country.
Rodrik (2005) calculates the social cost of holding reserves for developing
economies. Even when excluding the part of reserve accumulation due to the
traditional current account management purpose, his estimate of annual cost holding
reserves is around 1% of those countries’ GDPs. In Rodrik’s (2005: 9) words, “this
is a large number by any standard. It is a multiple of the budgetary cost of even the
most aggressive anti-poverty programs implemented in developing countries.
Mexico’s Progresa program, for example, cost around 0.2% of GDP.22”
22 Rodrik (2005) moreover argues that the excessive reserve accumulation in 1990s has not been rational. The emerging economies increased their short-term liabilities together with reserves. According to Rodrik, an optimal policy should have been reducing short-term liabilities while increasing reserves.
30
In the remainder of the thesis, I will show the effects outflows of resident capital and
reserve accumulation in emerging markets after liberalization. I will put forward that
with the emergence of a new financing pattern after liberalization, capital outflows
and reserve accumulation created new financing requirements for these countries
apart from their structural financing needs.
31
CHAPTER III
METHODOLOGY
In this chapter, I will present the methodology that will be applied in the following
chapters. The first section explains the balance of payments system and its
traditional representation. I decompose balance of payments and build new
components, which are explained in the second section. The third section is on data
and sample selection issues.
3.1. Balance of payments accounts and the traditional representation
As defined in IMF (1996:1) balance of payments is “a statistical statement that
systematically summarizes, for a specific time period, the economic transactions of
an economy with the rest of the world.” The balance of payments is concerned with
transactions and accordingly it deals with flows rather than stocks. These include
transactions in goods, services and income; transfers such as worker’s remittances,
and transactions related to acquisition or disposal of external financial assets.
32
The balance of payments statistics are collected by Central Banks in each country.
The national statistics are gathered by the IMF and published in its Balance of
Payments statistics and International Financial Statistics.
As balance of payments accounts reflect the economic transactions of an economy
with the rest of the world, transactions between residents of a country and non-
residents are tracked in these accounts23.
The double-entry bookkeeping system is employed in the balance of payment
accounts. In double-entry bookkeeping system, every transaction results in two
entries – one for the giving side, one for the receiving side. Credit items reflect
payments to residents by non-residents (inward the country) and debit items
reflecting payments from residents to non-residents (outward the country).
Examples of activities that give rise payments inward the country are exports, debt
issuance and inward foreign direct investment; while examples of activities that
give rise to payments outward are imports, investments by residents abroad and
divestments by non-residents by selling shares or closing down facilities.
Traditionally, credits are recorded with plus signs and debits are recorded with
minus signs. Hence, the total balance of payments account is zero.
23 Resident of a country must have a center of interest in that country. IMF (1995:13) states a unit has center of economic interest in a country “when there exists some location (dwelling, place of production, or other premises within the economic territory of the country) on, in, or from which the unit engages and intends to continue engaging (either indefinitely or over a finite but lengthy period of time) in economic activities and transactions on a significant scale.”
33
34
The standard components of balance of payments are current account and capital and
financial account. The transactions in goods, services, income and current transfers
are grouped under current account. The capital transfers and transactions in
country’s external financial assets and liabilities are collected under capital account.
The traditional representation is illustrated in Figure 3.
Figure 3. Traditional Representation of Balance of Payments Credit Debit
Trade in goods and services
Transfers
Income
Current account balance
Liabilities Claims
Direct investment by non-residents (to Turkey) Direct investment by residents (to abroad)
Portfolio investment bu non-residents (to Turkey)
• Equity investment
• Bond investment
Portfolio investment by residents (to abroad)
Other investment by non-residents (to Turkey) Other investment by residents (to abroad)
Capital account balance
Errors and omissions
Overall balance
Changes in reserve assets
0
It is worth presenting the recording mechanism of the balance of payments accounts
with an example. Consider a Turkish firm importing goods from the United
Kingdom for $10 million. The transaction is recorded as a debit item ($10M) in the
current account. There has to be balancing entry: If Turkish firm pays with a check
drawn on a Turkish bank, the corresponding transaction in the financial account is
recorded as an increase in Turkish liabilities to non-residents (a credit of $10M). If
the payment is drawn against an account the Turkish firm has in a British bank, the
corresponding transaction in the financial account is recorded as a reduction in
Turkish assets (a credit of $10M).
Normally, the total of current and capital and financial account balances should sum
up to zero. “In practice, however, when all actual entries are totaled, the resulting
balance will almost inevitably show a net credit or a net debit. That balance is the
result of errors and omissions in the compilation of statements.” (IMF, 1995: 38)
The difference between zero and sum of current and capital and financial accounts
forms the net errors and omissions item in the balance of payments24. With net errors
and omissions included, by construction, the three aggregates in balance of
payments sum up to zero.
The current account comprises of four basic items: trade in goods, trade in services,
income account and current transfers. The content of items in trade in goods and
services are obvious and necessitates no further explanation. The transactions
recorded in transfers are transactions “whenever an economy does not receive or
supply recompense—in the form of real resources or financial items—for goods,
services, income, or financial items supplied to or received from another economy.”
24 I will discuss to what extent the net errors and omissions are due to statistical approximations in the next section.
35
(IMF, 1996: 88) The current transfers include government grants and worker’s
remittances.
Items in income account, meanwhile, are earnings arising from the provision of the
factors of production. As land is associated with residence, the factors of production
whose earnings are recorded in balance of payments are labor and capital. Earning
on labor are compensations for employees. The income of capital can be divided into
two major groups: (1) the income from direct investment are the dividends that
investor earns from its direct investments and (2) the income from portfolio and
other investments, which includes dividend income earned from equities and interest
income from bond investments and trade and other credits.
The capital and financial account has two major sub-accounts, capital account and
financial account. Capital account includes capital transfers, i.e., transfer of
ownership of a fixed asset or the forgiveness, by mutual agreement between creditor
and debtor, of the debtor’s financial liability when no counterpart is received in
return by the creditor. Moreover, the country’s transactions with non-residents in
non-produced and nonfinancial assets (such as patents, copyrights, and licenses) are
also included in capital account.
The financial account, broadly speaking, keeps track of transactions of financial
assets. The credit items represent financial transactions that cause capital inflows to
the country; whereas the debit items represent financial transactions that cause
capital outflows. Functional classification of the items in financial account results
36
in four groups: direct investment, portfolio investment, other investment, and reserve
assets. The credit and debit items in the direct investment category represent
direction of investment: credit items are direct investments into the country by non-
residents; and debit items are direct investments by residents abroad. In the portfolio
and other investment categories the credit items are the transactions related to the
liabilities of residents to non-residents; and debit items are transactions related to the
assets of non-residents in the home country.
The direct investment category records transactions of investments “in which a
resident entity in one economy acquires a lasting interest in an enterprise resident in
another economy”, as defined by IMF (1995: 86). This long-lasting interest is
conventionally defined as a ten percent share or voting power. It should be noted
that not only equity capital and reinvested earnings are recorded in direct investment
account, but also the credits by the investor to the invested enterprise are recorded in
this category.
The portfolio investment category keeps record of transactions in equities (not
recorded under direct investment), other securities (debt instruments, i.e., bills,
bonds and notes) and financial derivatives. “The essential characteristic of
instruments classified as portfolio investment is that such instruments are traded or
tradable. That is, the instruments offer investors the flexibility to shift, regardless of
the underlying maturity of the instrument, invested capital from one instrument to
another. Portfolio investors are more concerned than direct investors about rates of
37
return that are independent of any influence investors may have and about being able
to move funds quickly if circumstances so dictate.” (IMF, 1996: 124).
Other investment is a residual category that includes all financial transactions not
considered direct investment, portfolio investment, or reserve assets. Other
investment category can be divided into four sub-categories: (1) trade credits, (2)
other loans, (3) currency and deposits, (4) use of at IMF credits and loans, and (5)
other assets and liabilities. Trade credits are to assets and liabilities that arise from
the direct extension, during the normal course of trading, of credit from a supplier to
a buyer. Loans are financial assets that are created through the lending of funds by a
creditor (lender) directly to a debtor (borrower); the lender receives no security
evidencing the transaction or receives a nonnegotiable document or instrument. The
currency and deposits include the currency issued by foreign governments and held
by residents represent claims that holders have on issuing governments and notes
and coins that are issued by the economy’s government and held by nonresidents
which represent an economy’s external liabilities. The use of IMF credits and loans
item is obvious and requires no further explanation.
The last functional category in the financial account is the reserve assets. They
consist of financial instruments available to the central authorities for financing or
absorbing an imbalance of payments or for regulating the size of such imbalances.
As debit items relate to assets held by residents, a debit item (a negative entry) in
38
reserve assets account represents an increase in reserves. Similarly, a credit item
shows a decrease.
3.2. A new structural decomposition
For my analytical purposes, I decompose the traditional components of the balance
of payments and following Laurent et al. (2002), I build new aggregates. The salient
feature of the new aggregates is the differential treatment of credit and debit entries
in the financial account. I demonstrate the new aggregates in Figure 4 and explain
them below.
Basic Balance (BB): Basic balance is defined as the current account balance plus
capital account and direct investment category of financial account. It therefore
gives a measure of country’s debt-financing requirement.
Trade in goods and services, income account, current and capital transfers, and net
foreign direct investments (FDI) are included in the basic balance. The capital
transfers listed in the capital account are in their nature no different than the current
transfers listed under current account, so consolidation of two accounts under one
title is reasonable.
Similarly, direct investment category of the financial account is different from the
other categories of the same account in two fundamental aspects: First, unlike most
39
other credit entries in the financial account (excluding equities in the portfolio
investment category), credit items in direct investment category do not represent
creation of debt to the rest of the world. Second, direct investment typically behaves
differently than the other categories. It is less volatile and generally driven by
explicit corporate strategies rather than short-term yield motivations or herd
behaviors. Therefore, direct investments are also treated differently from other items
in financial account and put into the basic balance component.
Figure 4. The structural components of the Balance of Payments Credits / Liabilities Debits / Claims
Basic Balance (BB)
Current Account Balance
Direct Investment, net
Capital Outflows (KO)
Portfolio investment
Other investment
Errors and omissions
Debt Flows (DF)
Portfolio investment
Other investment
Change in Reserves (R)
Source: Adapted from Laurent et al. (2002)
Debt Flows (DF): The transactions that increase country's gross liabilities to the rest
of the world are included in the debt flows component. These are the credit entries
in portfolio investment and other investment categories, i.e., the entries on the
40
credit side of the capital account (other than FDI) are collected in debt flows
aggregate.
The debt flows component provides a broad measure of the debt-generating capital
inflows to the country. However, there are two caveats: First, equity investments in
the portfolio investment category are in fact not debt-generating. Nevertheless, as a
general fact, international investors’ motive to engage in transactions with emerging
market equities is gaining short-term gains, not long-term dividends. In this respect,
investors’ motive to buy or sell equities is not different than buying or selling bonds
and other debt instruments in the portfolio investment category. Therefore, volatility
of equity investments is similar to the other portfolio investments and structurally
higher than direct investments. Lastly, data availability problems avoid a healthy
break-down of some countries portfolio investment category into equity and bond,
etc. sub-categories. Consequently, I opted for including equity investments into debt
flows component rather than basic balance25. Second, as stated in the previous
section, loans given by investors to their direct investment enterprises are included
in the direct investment category. These transactions are in reality debt creating.
However, because of practical reasons, in my analysis, these transactions are
included in basic balance rather than debt flows.
Capital Outflows (KO): These are gross claims of residents on non-residents
excluding FDI. These are the debit entries in the portfolio investment and other
25 In this respect, my structural decomposition is different from Laurent et al. (2002)
41
investment categories of the financial account. The value of the net errors and
omissions aggregate is regarded as a measure of “unrecorded capital flight” and
also included in the capital outflows component.
Theoretically, "errors and omissions" are due to accounting errors from the
approximation methods used. However, following Lane and Milesi-Ferretti
(2002) and Laurent et al. (2002), I treat net errors and omissions as
“unrecorded flight of resident capital” and hence include them in the capital
outflows component26.
The net errors and omissions measure net unrecorded transactions in the
balance of payments. As Lane and Milesi-Ferretti (2002: 3) state “this item
measures (net) unrecorded transactions that could reflect the mismeasurement
of transactions in current account or financial account or both. If it reflects
unrecorded trade transactions, we should adjust the current account
accordingly. If it reflects unrecorded financial account transactions, we should
add it to capital flows.” Therefore the first assumption when treating net errors
and omissions as unrecorded flight of resident capital is that the
mismeasurement is in financial transactions not in trade transactions. The
unrecorded financial transactions may either be capital inflows or outflows. So
I further assume that all unrecorded capital inflows are reductions in the stock
of assets held abroad by domestic residents.
26 Dooley (1988), Dooley and Kletzer (1994), Claessens and Naude (1993), UNCTAD (1999) and Boratav (2003) also treat net errors and omissions as unrecorded flight of resident capital.
42
Given the capital flight phenomenon in emerging market countries, the
assumptions seem reasonable. Laurent et al (2002: 12) argue that errors and
omissions seem to be larger in emerging market countries. “Sixty percent of
errors and omissions in the global balance of payments are due to the emerging
market countries, whereas their combined GDP accounts for just over 20% of
world GDP.” This may, of course, also be due to underdevelopment of these
countries statistical systems. But it is worth noting that the significant
improvement in data collections systems in the last 25 years has made no
significant effect in lowering the size of net errors and omissions. After
liberalization, possible lowering of uncontrolled capital movements is expected
to result in a decline in this item, which did not happen.
It should be noted that for some countries which liberalized trade of foreign
currency between residents, the net errors and omissions consist not only of
uncontrolled capital flight, but also to some extent foreign currency
movements from formal to informal economy. However, these movements are
similar to the unregulated capital flight both in their causes and consequences,
so that we may treat them together for our purposes.
Change in Reserves (R): This aggregate measures the change in country's
official reserves. As they are claims on the rest of the world, traditionally
reserves are placed in the financial account. However, the cost of holding
reserves as explained in Section 2.3 makes treatment of them under another
43
44
component informative. In accordance with the traditional representation, a
negative balance in this component represents an increase in official reserves. It is
worth noting that the reserves in an emerging market economy are not limited to
official reserves, banks and other resident agents also hold reserves, however
these reserves appear in other accounts and it is only possible to treat official
reserves as a separate component.
As balance of payments net out to zero, the new aggregates satisfy:
0=+++ tttt RKODFBB (1)
Dividing both sides of (1) to the previous years’ external debt stock, we obtain
the following:
01111
=+++−−−− t
t
t
t
t
t
t
t
DR
DKO
DDF
DBB (2)
1111 −−−−
−−−=t
t
t
t
t
t
t
t
DR
DKO
DBB
DDF (3)
The contributions of basic balance, capital outflows and change in reserves to the
change in the external debt stock at each year is demonstrated in (3). Ceteris
paribus, external debt stock will rise, when there is a basic balance deficit, i.e.,
0<tBB ; when capital outflows take place, i.e., 0<tKO ; and when official
reserves are accumulated, i.e., 0<tR .
45
Further decomposition of BB may be informative regarding the source of debt-
financing requirements. The income account in BB, which records net interest
and dividend payments, is characteristically negative for developing countries, as
they generally have the burden of past debt or they are recipients of foreign
investment. When this item is excluded, the remaining balance of BB will show
the financing requirement that is not a result of historical factors. Moreover, the
balance of income account is typically exogenous to policy decisions.
Nevertheless, there is more policy control on the remaining items: trade flows can
be adjusted using exchange rate adjustments or trade barriers and FDI can be
attracted by regulatory measures to attract foreign investors and privatizations.
Therefore, the remaining part of the BB from income account will be revealing
about the country’s efforts to correct disequilibrium in its balance of payments.
Consequently, I further decompose BB into net investment income (NI) and basic
balance excluding net investment income components:
ttt BENINIBB += (4)
Now putting (4) into (3):
11111 −−−−−
−−−−=t
t
t
t
t
t
t
t
t
t
DR
DKO
DNI
DBENI
DDF (5)
46
The interpretation of (5) is straightforward: Debt flows increase when there is a
deficit in income account, i.e., 0<tNI ; or when there is a deficit in basic balance
excluding income account, i.e., 0<tBENI .
For a large time span, if one wishes to explore what was financed with debt flows
then another equation may be useful. From (1):
tttt RKOBBDF −−−= (6)
1=−−−t
t
t
t
t
t
DFR
DFKO
DFBB (7)
Equation (7) decomposes debt flows in a given period according to their usage.
However one should be careful on the balance of debt flows. If debt flows gives
negative balance in a certain period, that is, the country has not been a debtor but a
creditor, than equation (7) will not be interpretative.
3.3. Data and sample selection
In my analysis, I calculate each of the new structural aggregate for each year for the
countries in my sample. The data for the components come from the balance of
payments statistics reported in IMF International Financial Statistics (IFS). For the
calculation of equation (3) and (5) I use external debt stock data compiled by the
World Bank and reported in the World Development Indicators (WDI).
47
In order to make comments on the relationships given in equations (3) and (5), a
look at the relationship between debt flows aggregate and external debt stock is
illustrative. The total of cumulative debt flows to the countries in my sample and
total external debt stocks of these countries are presented in Figure 5. Apparently,
the two series move with the same pattern. This relationship justifies the observation
that 1−t
tD
DF provides a measure of the percentage increase in debt stock in the
current year.
Figure 5. Debt flows and external debt stock for the sample countries
0
200
400
600
800
1,000
1,200
1,400
1,600
1975
1977
1979
1981
1983
1985
1987
1989
1991
1993
1995
1997
1999
2001
Debt flows (cumulative)External debt stock
Source: IFS, WDI; figures are in billion US dollars.
The differences between two series stem from various reasons: (1) there exists an
initial debt stock, (2) the volatility in exchange rate changes the total value of the
debt stock since debt is denominated in various currencies, (3) the interest accrued
but not paid is added to the debt, but it does not appear in the balance of payments,
(4) the two series are generated from different data sets by different institutions and
naturally there exists difference between them.
The list of the countries that I will analyze is provided in Table 1. They can be
grouped into three geographical categories: Asia, Latin America and East Europe.
Several emerging market countries are excluded from my analysis. Among them,
China and India are significantly larger than the sample countries and their approach
to financial liberalization have been cautious. Singapore and Hong Kong have early
liberalization stories, however as they are very small by size if not by the size of
their financial markets and capital flows to these countries have different motives
(they serve as hubs for financial capital) than the countries in the sample. Russia is
again very large and its economy mostly depends on natural resources and as a result
a very different balance of payments picture emerges.
Table 1. List of Sample Countries
Asia Latin America East Europe
Korea Malaysia Indonesia Thailand Philippines
Argentina Brazil Chile Mexico Peru Columbia Venezuela
Czech Republic Hungary Poland Turkey
48
49
The data for all the countries in the sample is not perfect. For transition countries,
the data begins in late 1980s or early 1990s. For some periods in some countries in
the sample, there exist missing data for some forms of capital flows. It is impossible
to verify whether data is missing or it is equal to zero. However, I believe this
missing data issues does not create a bias in the country analysis.
For each country, and for regional and general totals, I engage in three exercises.
First, using equation (5), I graph the contributions of the new components of balance
of payments (BOP), that is, net investment income deficit (1−
−t
tD
NI ), basic
balance excluding investment income deficit (1−
−t
tD
BENI ), capital outflows
(1−
−t
tD
KO ) and reserve accumulation (1−
−t
tD
R ) to increases in external debt
stock (1−t
tD
DF ) (hereafter Exercise 1). Second, when it provides a useful
interpretation, I use equation (7) to decompose usage of debt flows (hereafter
Exercise 2). Third, where data are available and results are interpretative, I provide
correlations between the components in equation (1) for different periods of a
country (hereafter Exercise 3). I use quarterly data in Exercise 3.
As it is a point where data is available for most countries, the country tables start
from the year 1984 and cover the period until 2003. Annual data is used when
building the country tables. In the correlation analysis, quarterly data is used. For
some countries and some periods quarterly data is not available while annual data is.
I begin the correlation period from the earliest date where quarterly data is available
for each country and cover until 2003.
50
CHAPTER IV
LIBERALIZATION EXPERIENCES
This chapter contains the core analysis in my thesis, which uses the methodology
presented in Chapter Three. In the first section, I make an aggregate analysis for all
emerging markets and for Latin America, Asia and East Europe regions. In the
second section I undertake individual studies for 15 emerging market countries. In
the third section, I outline some similarities and differences in the 15 countries
analyzed and make a synthesis of the liberalization experiences.
4.1. Some stylized facts
In order to provide a general picture of contribution of our new aggregates to
increases in debt stock, I present results of Exercise 1 for the total of all countries in
my sample in Figure 6. The surge in debt flows after 1989 is apparent in the figure.
In some years, the emerging markets in the sample have got new total debt more
than a quarter of their previous year’s total debt stock. 1997 marks an end to the
surge. Beginning with the Asian crisis in 1997, a series of emerging market crisis
51
have stopped debt flows to these countries. For the sample countries, debt flows
fluctuate around zero after 1997.
Concerning the other components, investment income account has consistently gave
deficits, while the other components of the basic balance were either in surplus or
gave very small deficits. When, debt flows were abundant, the emerging market
countries did not gave surpluses in the other components of basic balance. When
debt flows stopped, they offset the investment income deficit with surpluses in the
other components. It can be argued that, when the emerging market countries in the
sample are treated in aggregate, investment income is the primary cause of basic
balance deficits, when they are treated as a whole.
Figure 6. Ratio of BOP components to external debt stock: All countries
-0.30
-0.20
-0.10
0.00
0.10
0.20
0.30
0.40
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
NIBENIKORDF
Source: author’s calculations, IFS
52
Capital outflows have consistently given positive balance after 1990. The capital
outflows are consistently larger in 1990s when compared to 1980s. After
liberalization, resident capital flight became a structural phenomenon in the sample
countries.
Meanwhile, reserve accumulation accompanied the debt flows surge after 1990. In
some years, reserves of up to 10% of previous year’s debt stock were accumulated.
The only negative balance in reserve accumulation is in 1997, where Asian
economies lost a significant portion of their reserves in their attempts to protect
exchange rate pegs.
Figure 7 shows debt flows to the sample countries in Asia, Latin America and
Eastern Europe. Debt flows to Asia began to rise in late 1980s and reached its peak
in 1996. Asian crisis of 1997 resulted in a sharp decline in debt flows to the region,
and after 2000 a small recovery is observed. For Latin America, 1980s began with a
debt crisis and a fall in debt flows. After negative balances in debt flows in 1980s,
there has been an impressive surge in 1990s, which sustained until 1999. With
contagion from Asian crisis, beginning with the Brazilian devaluation of 1999, a
significant reduction occurred. After the Argentine default of 2001, debt flows began
to give negative balance. Debt flows to East Asian countries was nil before 1990,
when most countries in the sample was under communist rule, and Turkey had a
financially closed economy. A surge similar to the other regions happened after
53
1990 when the countries acquired open capital account regimes. A fall in debt flows
is observed after 2000.
Figure 7. Debt flows to Asia, Latin America and East Europe (bil. $)
-60
-40
-20
0
20
40
60
80
100
1975
1977
1979
1981
1983
1985
1987
1989
1991
1993
1995
1997
1999
2001
2003
Latin AmericaAsiaEast Europe
Source: IFS
Data for several East European countries are not available prior to 1990. In these years, figures represent debt flows to countries whose data are available.
When the behaviors of our new aggregates are concerned, the contrasting positions
of Asia and Latin America in some components are impressive. The behaviors are
illustrated in the results of repetition of Exercise 1 for Latin America (Figure 8) and
Asia (Figure 9)27.
27 The inter-regional capital flows are also included in the calculations. For instance, any flow from Korea to Malaysia is included both in the KO component (for Korea) and in DF component (for Malaysia).
54
First, NI is in negative balance in all periods in both regions; however the deficit has
consistently been larger for Latin American countries. The legacy of past debt is the
major reason. External debt stock to GDP ratios are given for sample countries in
two regions in Table 2 In 1980, it was 36% for Latin America and 15% for Asia.
Larger debt resulted in larger interest payments to the rest of the world. Moreover,
rolling over the debt has been more costly for Latin America countries compared to
Asian countries, because of the short term structure of the debt. Partly because of
this reason, although the external debt stock to GDP ratios converged in 1990s, NI
continued to give larger deficits in Latin America. Lastly, large scale privatizations
to attract FDI has resulted in larger dividend payments to the rest of the world in
Latin America, especially in some countries such as Argentina, further aggravating
the imbalance in NI.
Second, Latin America gave consistent surpluses in BENI, whereas BENI surpluses
were nearly not present at all in Asia for the period of 1990 – 97. An interpretation
may be the necessity of giving trade surpluses in order to cover large NI deficits in
Latin America resulted in this pattern. FDI attracted to large scale privatizations also
had favorable effect on BENI surpluses while causing NI imbalances in turn.
Table 2. External Debt Stock / GDP ratios
1980 1985 1990 1995 2000 Latin America 36% 59% 39% 37% 40% Asia 15% 27% 27% 33% 44%
Source: WDI
55
Figure 8. Ratio of BOP components to external debt stock: Latin America
-0.30
-0.20
-0.10
0.00
0.10
0.20
0.30
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
NI
BENI
R
KO
DF
Source: author’s calculations, IFS
Figure 9. Ratio of BOP components to external debt stock: Asia
-0.30
-0.20
-0.10
0.00
0.10
0.20
0.30
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
NI
BENI
R
KO
DF
Source: author’s calculations, IFS
56
Table 3. Usage of Debt Flows in Sample Countries (1990 – 1996)
Total Asia Latin America East Europe Basic balance deficit 37% 44% 30% 37%Outflow of resident capital 34% 25% 46% 15%Reserve accumulation 29% 30% 24% 48%
Source: author’s calculations, IFS
Third, when the patterns of usage of debt flows are concerned, capital outflows
dominates other aggregates in Latin America, whereas basic balance finance was
more important than items in Asia. This phenomenon is obvious from the results of
the Exercise 2 separately carried out for Latin America and Asia and illustrated in
Table 3. The figures represent usage of debt flows between 1990 – 1996 in two
regions. Such illustration is not interpretative for alternative periods, since debt
flows give negative or very small positive balances. For the period of 1990 – 96,
nearly half of the debt flows financed capital outflows in sample countries in Latin
America. This ratio is a quarter for Asia, while basic balance deficit has a share of
44%.
Figure 10 demonstrates the contributions of the aggregates to increase in debt stock
for East European countries in the sample. As the sample size is relatively small, the
behavior of the aggregate values is partly dominated by Turkish figures, evident
from negative values of DF in the crisis years of 1994 and 2001. After liberalization
in 1990s, the surge in debt flows mostly financed reserve accumulation. NI gave
deficit balances, albeit in low values. BENI has characteristically been in surplus
57
values except 1993. Capital outflows, although much lower than reserve
accumulation has also been a major item financed through debt flows.
Figure 10. Ratio of BOP components to external debt stock: East Europe
-0.20
-0.15
-0.10
-0.05
0.00
0.05
0.10
0.15
0.20
0.25
1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003
RKONIBENIDF
Source: author’s calculations, IFS
4.2. Country analysis28
In this section, I will make brief analyses of balance-of-payments components for
each country in the sample. The analysis will be based on results of Exercise 1, i.e.,
figures representing contributions of our new components to external debt stocks of
28 Readers who are not interested in country details may skip this section without loss of continuity. The next section summarizes the country experiences.
58
the countries. For the periods where debt flows are significantly positive, I also
analyze their usage by carrying out Exercise 229. I do Exercise 3 for Argentina,
Mexico, Chile, Brazil, Korea, and Philippines30. A brief history of liberalization
experiences for each country, together with references for historical explanations, is
provided in the Appendix.
4.2.1. Argentina
The decade of 1980s in Argentina was characterized by chronic high inflation and
economic stagnation. An early capital account liberalization attempt in 1977, which
was implemented without trade liberalization resulted in a significant capital flight.
The liberalization decision was reversed in 1982 and Argentina repressed its
financial system in 1980s. Argentina’s capital account liberalization was a part of a
large-scale reform package introduced by newly elected Menem administration in
1991 to achieve macroeconomic stability. Most important element of the package
was the “Convertibility Plan,” a currency board regime where the Argentine peso
was tied to US dollar at one-to one parity. The package also included large scale
privatizations to attract FDI. Capital account liberalization was one of the most
29 I do not undertake Exercise 2 in some countries: In Venezuela, Malaysia, and Czech Republic, the periods that debt flows are consistently positive are too short for a meaningful decomposition. In Peru, capital outflows are negative. In some of the the countries which I do Exercise 2, pre-liberalization debt flows are negative rendering a comparison between pre- and post- liberalization periods impossible. 30 For other countries either quarterly data are not available for pre-liberalization and post-liberalization periods and / or it is impossible to divide the period analyzed into different phases with meaningful differences in correlation values.
59
complete ones at that time, by easing all restrictions on capital inflows, outflows and
allowing residents to hold foreign currency deposit accounts in domestic banks.
Moreover, dollarization has been encouraged by the administration by making it
legal to write contracts in foreign currencies.
Figure 11. Ratio of BOP components to external debt stock: Argentina
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
0.4
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
NIBENIRKODF
liberalization
crisis
Source: author’s calculations, IFS
Figure 11 demonstrates results of Exercise 1 for Argentina. A surge is observed in
debt flows in 1991. The debt flows have consistently been higher after 1991, with a
peak in 1993 at a level more than one third of the external debt stock. The contagion
from Mexican tequila crisis has been very limited. Argentina had an external debt
stock at the level of 85% of GDP prior to liberalization and had given negative
60
balance thought 1990s. FDI inflows have contributed to BENI, which has surpluses
in some years, nevertheless, profit transfers from FDI investments further augmented
NI deficit in late 1990s. Capital outflows have been encouraged by the
“Convertibility Plan,” as many contracts were denominated in foreign currencies,
residents hedged themselves by acquiring foreign assets.
Meanwhile, inflation was virtually eliminated in mid-1990s. However, the relative
macroeconomic stability was undermined by two major risk factors: (1) the
overvaluation in peso which was tied to US dollar; (2) fiscal deficits of federal states
which had constitutional right to issue their own debt instruments independent from
the central government. The reduction in the capital flows to emerging markets after
Asian crisis of 1997 has made these risks harder to manage. When Brazil devalued
real in 1999, it has been a real loss of competitiveness for Argentina and the
overvaluation in Argentine peso has become a serious impediment on economic
activity. Given international investors’ risk avoidance, debt flows have been reduced
in 1999, while outflows of resident capital continued. In 2001, the capital flows were
reversed, as international investors started to pull their money back from Argentine
assets. A depletion of reserves made it impossible to defend the peso peg any more,
and the currency board collapsed in December 2001.
The structural change in the pattern of debt flows and their usage after liberalization
in Argentina is obvious. Before liberalization debt flows has been nil or negative,
and they were associated with basic balance finance. Capital outflows were not
61
apparent. After liberalization, debt flows has mostly been associated with capital
outflows. Correlation between capital outflows and debt flows, as demonstrated in
Table 4, rises from 0.31 in 1980 – 1990 period to 0.76 in 1990 – 1998 period; while
correlation between debt flows and basic balance deficit falls from 0.46 to 0.36.
Reserve accumulations role in the usage of debt flows is limited although Argentina
had a currency board regime. The decomposition of debt flows usage between 1990
and 1998, as seen in Table 5, shows that more than half of debt flows financed
capital outflows. A third of debt flows were used to finance basic balance deficits
and the remaining 20% was used to accumulate reserves.
Table 4. Correlation of DF with other components in Argentina
1980 - 1990 1990 – 1998 BB 0.46 0.36 BB(-1) 0.40 0.15 BB(+1) 0.46 0.27 KO 0.31 0.76 R 0.37 0.26 Source: author’s calculations, IFS Table 5. Usage of Debt Flows in Argentina (1990 – 1998)
Basic balance deficit 31%Outflow of resident capital 50%Reserve accumulation 19%
Source: author’s calculations, IFS
4.2.2. Mexico
Mexico has begun capital account liberalization in 1989; and by 1992 international
financial liberalization has been completed in Mexico. The behavior of components
62
of balance of payments can be analyzed in three phases. The first phase covers the
pre-liberalization period. The second phase begins in 1990 and ends with the Tequila
crisis of 1994. The year 1994 witnessed a reversal in debt flows. The third phase
begins in 1995 and spans the years afterwards.
Figure 12. Ratio of BOP components to external debt stock: Mexico
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
0.4
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
NIBENIRKODF
liberalizationtequila crisis
I II III
Source: author’s calculations, IFS
Figure 12 shows the results of Exercise 1 for Mexico. Debt flows fluctuate around
zero in the pre-liberalization phase. The second phase is characterized by a high
level of debt flows, reaching 30% of the external debt stock at its peak in 1993. The
period also witnessed a reduction in inflation and restoration of fiscal balances.
During this period, not only NI, but also BENI account has been in deficit. Major
63
source of finance need has been this basic balance deficit. Moreover, as the domestic
investors used the opportunity provided by capital account liberalization to hedge
themselves given the memory of high inflation in 1980s, the period has witnessed a
consistent flight of resident capital. Reserve accumulation also had a role, albeit
lower than the second phase, in the usage of debt flows. The financing need that
stemmed from basic balance deficits and capital outflows culminated in the first
emerging market crisis of the decade when debt flows reversed in 1994.
The third phase is characterized by a lower level of debt flows. Although NI deficit
lasted in the second phase, BENI deficit was eliminated and BENI gave surpluses
for most of the years. Annexation to NAFTA has provided an anchor for relative
stability in the domestic economy. Capital outflows were reduced. In the third phase,
reserve accumulation has been relatively higher than the second phase.
Table 6. Correlation of DF with other components in Mexico 1980 - 1989 1990 – 1993 1995 - 2003 BB 0.68 -0.10 -0.14 BB(-1) 0.67 0.08 -0.05 BB(+1) 0.60 0.36 -0.14 KO 0.61 0.92 0.64 R 0.10 0.34 0.35 Source: author’s calculations, IFS Table 6 demonstrates correlations between debt flows and other components. The
differences between pre-liberalization period and the two phases after liberalization
are striking. Before liberalization, the debt flows were mostly associated with basic
64
balance deficits with a correlation of 0.68. After liberalization, although basic
balance continued to give large deficits, the correlation between debt flows and basic
balance deficits has virtually been eliminated. Rather, debt flows has been associated
with capital outflows, especially in the second phase. Their correlation in the second
phase is 0.92. Correlation between debt flows and reserve accumulation also rose,
albeit to a lower extent, after liberalization.
Table 7. Usage of Debt Flows in Mexico (1990 – 1994) 1990 - 1993 1990 - 1994Basic balance deficit 63% 75%Outflow of resident capital 15% 23%Reserve accumulation 22% 2%
Source: author’s calculations, IFS
As the second phase is the only period when debt flows have given a significant
positive balance, it is meaningful to engage in the analysis of usage of these flows in
this period. The analysis presented in Table 7 shows that basic balance deficits have
been the foremost important item to be financed in this period. Capital outflows and
reserve accumulation are head to head, subject to the inclusion of 1994 during which
reserves were depleted.
4.2.3. Chile
Chile had one of the most stable macroeconomic climates in Latin America in
1990s. This climate is partially a result of reduction in public debt after the debt
65
crisis of 1980s with prudent fiscal policies. Moreover, Chile’s approach to capital
account liberalization has been differentiated by most Latin American countries with
the controls on capital inflows, encaje.
The capital account liberalization in Chile started in 1985 and gradually most
restrictions were eliminated in the late 1980s. In 1991, Chilean authorities
implemented unremunerated reserve requirements, encaje, on capital inflows.
Figure 13. Ratio of BOP components to external debt stock: Chile
-0.4
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
0.4
0.5
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
NIBENIRKODF
liberalization
Source: author’s calculations, IFS
Encaje was an implicit tax on inflows as a certain portion (30% in the beginning) of
capital inflows were obliged to be deposited at a non-interest bearing account in the
Central Bank for a year. The role of encaje in limiting inflows is very much debated
in the literature. Although its role in limiting inflows is not obvious, there is almost a
66
consensus in literature about its success in lengthening the maturity of debt flows.
Figure 13 reveals results of Exercise 1 for Chile. The debt flows has consistently
been positive in Chile in 1990s and contagion from all emerging market crisis,
including neighbor Argentina, has been very limited. With commitment to attract
FDI and large-scale privatizations and deregulations to this end, Chile has attracted
large amounts of FDI. The NI deficit and BENI surpluses has been head-to-head,
rendering almost zero basic balance in 1990s. Outflows of resident capital have been
limited until 1998. Meanwhile, as a sign of a stable macroeconomic climate, Bank of
Chile accumulated a large amount of reserves in 1990s.
Chilean authorities have virtually eliminated encaje in 1998. Partially because of this
policy-change, debt flows have not been reduced sharply as it had been in many
emerging markets at the end of 1990s. The large capital flight figures at the end of
1990s are due to the ease of restrictions on pension funds in acquiring foreign assets
in this period. As the pension system is privatized in Chile, these funds administer a
large portion of domestic savings and ease of restrictions has resulted in a one-shot
surge in outflows. Nevertheless, stable macroeconomic climate seems to have
avoided large scale outflows of resident capital in Chile.
Correlations presented in Table 8 reveal that most of the debt flows have been
associated with reserve accumulation in 1990s. Correlation between debt flows and
capital outflows rises after 1998. There is almost no meaningful association between
debt flows and basic balance. Indeed, as seen in Table 9, more than 90% of debt
67
flows were used to accumulate reserves between 1991 – 1997. The remaining
portion had been used to finance outflows and portion used for basic balance finance
is nil.
Table 8. Correlation of DF with other components in Chile
1991 - 1997 2000 – 2003 1991 - 2003 BB 0.03 -0.23 -0.08 BB(-1) 0.13 -0.18 0.00 BB(+1) -0.04 0.22 0.15 KO 0.44 0.54 0.39 R 0.63 0.55 0.36 Source: author’s calculations, IFS
Table 9. Usage of Debt Flows in Chile (1991 – 1997)
Basic balance deficit 1%Outflow of resident capital 8%Reserve accumulation 92%
Source: author’s calculations, IFS
4.2.4. Brazil
Brazil’s position regarding international financial liberalization has been more
hesitant compared to its Latin American neighbors. Rather than implementing full
capital account liberalization a la Argentina and Mexico or systematic controls on
inflows a la Chile; Brazil opted for a modest opening of capital account with
sporadic applications of capital controls. The liberalization degree varied according
to domestic economic and political situation and international financial climate.
68
Nevertheless, it has been limited at all. For instance, it is still forbidden for residents
to hold foreign currency deposits in domestic banks.
Brazil began 1980s with a debt crisis. Debt flows have been interrupted during this
decade. Macroeconomic instability with three-digit inflation rates and public deficits
over 10% as a ratio to GDP has been the major characteristic of 1980s. Brazil took
first capital account liberalization steps in 1991 in this environment of
macroeconomic stability. Debt flows immediately rose as a response to
liberalization. Real plan, a macroeconomic policy package including an exchange
rate peg has been successful in substantially reducing inflation. Restrictions applied
in 1993 did result in a fall of debt flows as international investor’s confidence in
Real plan caused them to lend more. The controls were eased in 1995 to avoid
contagion from Tequila crisis. Indeed, contagion effect has been very limited. In
1996, controls were applied again. Asian and Russian crises in 1997 and 1998 have
affected capital flows to Brazil, lowering debt flows component significantly in
these years. A subsequent ease of controls to attract more inflows was not
successful. As reserves began to deplete and it has become impossible to sustain
external balances, Brazil devalued real in 1999. After 1999, although non-residents
are allowed to invest under the same rules with residents, debt flows have been very
low.
Results of Exercise 1 for Brazil can be found in Figure 14. Three phases can be
identified in the behavioral structure of balance of payments components during the
69
capital account liberalization experience of Brazil. The first phase covers the pre-
liberalization period until 1991. The second phase spans the period of 1991 – 1999.
The third phase spans the period after the devaluation of 1999.
In all phases, reflecting the debt burden of country since 1980s, NI has been in
negative balance. In the first phase, debt flows are negative. The second phase is
characterized by a positive balance of debt flows each year, reaching its peak in
1995 around 20% of the external debt stock. It should be noted that this figure is
lower than some Latin American countries such as Mexico and Argentina. The
BENI balance has been positive for the first three years of the first phase; however,
surplus became very small or even negative after 1995. Therefore it offset the NI
deficit only in the first years. Capital flight consistently rose after 1991, stopped in
1995 when the confidence was restored through Real plan, and then continued.
Reserve accumulation has been high until 1996, but much of the accumulated
reserves have been lost in order to defend the peg of real at the end of the first phase.
According to the figures in Table 10, 44% of debt flows in this period has been used
to finance reserve accumulation, 33% to finance capital outflows and the remaining
22% to finance basic balance deficits. If the analysis period is limited to 1991 – 96,
in order to omit the effect of reserve depletion after 1997, the share used to finance
reserves rises to 56%. In the third phase, as debt flows stopped, BENI adjusted to
cover the deficit in NI. Capital flight continued, while reserve accumulation has been
very limited.
70
Table 10. Usage of Debt Flows in Brazil (1992 - 1998)
1992 – 1998 1992 - 1996Basic balance deficit 23% 24%Outflow of resident capital 44% 56%Reserve accumulation 33% 20%
Source: author’s calculations, IFS
Table 11. Correlation of DF with other components in Brazil
1975 - 1991 1992 - 1998 1999 – 2003 BB 0.52 0.08 0.29 BB(-1) 0.55 0.32 -0.06 BB(+1) 0.33 0.00 -0.11 KO 0.23 0.29 0.18 R 0.24 0.91 0.88 Source: author’s calculations, IFS
Figure 14. Ratio of BOP components to external debt stock: Brazil
-0.25
-0.2
-0.15
-0.1
-0.05
0
0.05
0.1
0.15
0.2
0.25
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
NIBENIRKODF
devaluation of real
liberalization
Real planI II III
Source: author’s calculations, IFS
71
The break between the relationship of basic balance finance and debt flows after
liberalization is evident from correlations presented in Table 11 The correlation
between two components in the pre-liberalization period was 0.55. After
liberalization, in the second phase it fell to 0.08. Meanwhile, correlation between
reserve accumulation and debt flows rose from 0.24 to 0.91. Association between
debt flows and capital outflows has been relatively stable. The analysis of the
second phase, when debt flows were abundant, reveals that debt flows became a
mean to finance reserve accumulation and debt flows rather than basic balance
deficits after liberalization.
4.2.5. Peru
Peru is one of the members of the small set of emerging market countries which
have successful liberalization experiences. It witnesses recurring balance of
payments crisis in 1970s and early 1980s; high chronic inflation lasted until early
1990s. In 1990, Peru implemented an economic reform package which includes
capital account liberalization. Other elements were restoring of fiscal balances,
strong bank supervision and floating exchange rates. In 1994, differential treatment
of residents and non-residents were eliminated.
Figure 15 demonstrates results of Exercise 1 for Peru. Debt flows began to give
positive balance after 1990, and especially between 1994 – 1999, they have been
72
consistently positive. However, the magnitude of debt flows has been smaller as a
ratio of external debt stock compared to other emerging market countries. Reflecting
the debt burden of Peru, NI had always been in large deficit. BENI had been in
deficit before 1994, and then it gave surpluses, albeit not large enough to offset NI
deficits. Basic balance deficit had been the major financing need prior to 1998. It
had been larger than debt flows and been offset by negative balance in capital
outflows, i.e., residents switched to domestic assets to finance basic balance deficits.
Capital flight has been small also after 1998. Reserve accumulation has also been
limited, even negative before 1994.
Figure 15. Ratio of BOP components to external debt stock: Peru
-0.3
-0.25
-0.2
-0.15
-0.1
-0.05
0
0.05
0.1
0.15
0.2
0.25
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
NIBENIRKODF
liberalization
Source: author’s calculations, IFS
73
Debt flows to Peru mostly financed its income account deficits and to a smaller
extent reserve accumulation. As capital flight has been low, even negative, the
financing need has been small. The relatively low level of debt flows combined with
prudent policies avoided large reversals and financial crisis.
4.2.6. Columbia
Columbia’s Apertura Program implemented in 1991 included international financial
liberalization along with other standard measures of such economic reform
programs. However, Columbia also used Chilean style unremunerated reserve
requirements between 1992 – 2000. The rate of the URRs was raised after 1994 and
substantially lowered in 1998. Nevertheless, the effects of these controls are mixed
in limiting debt flows as it is the case in Chile.
Figure 16 displays results of Exercise 1 for Columbia. NI is consistently in deficit
after 1990. BENI has also been in deficit for many years before 1999. Consequently,
more than a half of debt flows are used for basic balance deficit finance between
1993 - 1999. This can be observed from Table 12, which reveals financing patterns
in Columbia. The remaining part mostly financed capital flight (41%) and to a small
extent reserve accumulation. After 1999, debt flows are reduced and they have small
negative balances in some years. The fall in debt flows was offset by BENI
adjustment to give positive balances. Columbia did not manage to eliminate inflation
and fiscal deficits in 1990s. The NI deficit continued to constitute a financing need
74
after 1999; however BENI adjustment and small scale of debt flows reversal avoided
a large scale financial crisis.
Figure 16. Ratio of BOP components to external debt stock: Columbia
-0.4
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
1984 1986 1988 1990 1992 1994 1996 1998 2000 2002
NIBENIRKODF
liberalization
Source: author’s calculations, IFS
Table 12. Usage of Debt Flows in Columbia (1993 - 1999)
Basic balance deficit 55%Outflow of resident capital 41%Reserve accumulation 5%
Source: author’s calculations, IFS
4.2.7. Venezuela
Venezuela’s capital account liberalization history is mixed. 1990s witnessed large-
scale liberalization attempts and implementation of sharp restrictions. As an oil
75
exporter country, Venezuela consistently gives positive current account balances.
These two properties make Venezuela a special case among the emerging market
countries in the sample.
Figure 17 demonstrates results of Exercise 1 for Venezuela. The first phase of
liberalization is between 1989 and 1994. A small surge in debt flows happened in
this period. Extensive controls on both inflows and outflows were implemented
between 1994 – 1996, when parallel market premium in exchange rates reached 40 –
100%. Debt flows reversed in this period; however, it is not possible to observe the
same reversal in capital outflows. Rather they increased and stayed high when
capital account has been liberalized with IMF supported stand-by program in 1996.
Surge in debt flows was smaller in this liberalization period and ended in 2000 in
accordance with global reversal of capital flows to emerging markets. Capital
outflows stayed in deficit. Capital account is again repressed in 2003.
In Venezuelan experience, large BENI surpluses stemming from oil revenues offset
NI deficits and resulted in large positive basic balances. Debt flows have been very
small compared to basic balance surpluses. Nevertheless, in Venezuela, major
financing need comes from capital flight, which is consistently large even in
repression periods. The lack of confidence in domestic institutions and frequent
regime changes may be root causes of this behavior. Venezuelan experience shows
that given a certain level of financial development and initial liberalization, it is hard
to limit outflows by repressions in certain periods.
76
Figure 17. Ratio of BOP components to external debt stock: Venezuella
-0.6
-0.5
-0.4
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
0.4
0.5
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
NIBENIRKODF
liberalizationcapital controls
liberalization back
Source: author’s calculations, IFS
4.2.8. Korea
Korea has been one of the most rapidly growing countries in the world in 1980s.
This growth was fulfilled with a relatively closed financial system. Directed lending
has been used as an active instrument of industrial policy. Foreign direct investment
was also restricted. The industrial scene was dominated by chaebols, large
conglomerate firms supported and protected by the state.
Figure 18 demonstrates results of Exercise 1 for Korea. First minor capital account
liberalization attempts have been undertaken in late 1980s. Portfolio and other
77
investment inflows were liberalized rapidly between 1991 – 93, while controls on
outflows remained. Korea also has been reluctant to liberalize direct investments;
foreign ownership in Korean firms was still limited to 12% in 1995. Outflow
liberalization has been gradual and international financial liberalization was
completed with the OECD membership of Korea in 1996.
Debt flows poured into Korea after 1990 with rapid liberalization of inflows. Unlike
Latin American markets, low debt burden of Korea prior to liberalization has
resulted in very low NI deficits during this period. However, as a result of its
industrial structure, large outward direct investments by chaebols and currency peg
to US dollar, Korea gave consistent current account deficits. As foreign direct
investments were restricted, current account deficits translated into large and
negative BENI numbers. Therefore, low NI values did not avoid basic balance
deficits in Korea.
Although restrictions on outflows were eased in a gradual fashion, capital flight has
been very rapid especially after 1993. The period until 1997 was characterized by a
relatively low level of reserve accumulation. The pour of debt flows ended in 1997
with Asian crisis. After the crisis, debt flows reduced sharply. BENI has adjusted to
this reduction, as devaluation induced current account surpluses. Capital outflows
also fell, as more domestic capital was needed when external debt flows are not
available. Meanwhile, reserve accumulation gained dominance after 1997, probably
as a result of the lessons taken from the crisis.
78
Figure 18. Ratio of BOP components to external debt stock: Korea
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
NIBENIRKODF
gradual liberalization
Asian crisis
Source: author’s calculations, IFS
Korea’s capital account liberalization has the effect on association of debt flows with
other components of balance of payments in two stages. In the first stage, from 1990
to 1997, debt flows association with basic balance deficits prior to liberalization
persisted, but debt flows also been associated with capital outflows and reserve
accumulation. After Asian crisis, in the period of 1999 – 2003, as basic balance
deficits were eliminated, the association between basic balance and debt flows
reduced significantly. Correlation between debt flows and reserve accumulation
increased, and association of debt flows with capital outflows also continued. These
relationship patterns can be observed from the correlation figures in Table 13.
79
Table 13. Correlation of DF with other components in Korea
1977 - 1989 1990 – 1996 1999 – 2003 BB 0.81 0.71 0.25 BB(-1) 0.72 0.72 0.22 BB(+1) 0.75 0.77 -0.04 KO -0.04 0.86 0.62 R 0.04 0.44 0.66 Source: author’s calculations, IFS
Table 14. Usage of Debt Flows in Korea (1990 – 1996)
Basic balance deficit 42%Outflow of resident capital 46%Reserve accumulation 13%
Source: author’s calculations, IFS
Korea’s usage of debt flows, when they were in very high values in the period of
1990 – 97 are shown in Table 14. Consistent with the correlation figures, Korea has
used almost half of the debt flows to finance outflow of resident capital. 42% of debt
flows financed basic balance deficits. The portion of debt flows used to finance
reserve accumulation is relatively low at 13%.
4.2.9. Malaysia
Malaysia has been an early liberalizer of the capital account. International financial
liberalization had mostly been achieved in late 1970s and further restrictions were
eliminated in mid-1980s. Nevertheless, early liberalization did not bring an early
surge in debt flows.
80
Results of Exercise 1 can be found in Figure 19 for Malaysia. The surge in debt flow
to Malaysia happened parallel to surge in other emerging market countries, after
1990. Authorities responded this surge, when debt flows peaked at more than 30%
of the external debt stock in 1993, with imposition of temporary controls on inflows
in 1994. They were abolished in 1995. Indeed effect of capital controls had been
temporary, only a year-long fall in debt flows in 1994. However, recovery has not
been significant and debt flows have been very small until 1997. After Asian crisis,
they become negative. Malaysia responded the crisis with imposition of controls on
capital outflows which lasted from 1998 to 1999.
Figure 19. Ratio of BOP components to external debt stock: Malaysia
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
NIBENIRKODF
Asian crisis
liberal capital account
Source: author’s calculations, IFS
81
The total of debt flows in the period of 1990 - 96 is very small making an analysis of
their usage meaningless. Yet, some characteristic regarding the patterns of the
components of balance of payments are worth examining. The NI component gave
consistent negative balances, which were generally offset by positive value of BENI.
In a different pattern form many of the other emerging market countries, 1990s did
not witness capital flight in Malaysia. The value of outflows of resident capital was
negative in many years. On the other hand, parallel to other emerging markets, even
more than them, Malaysia used debt flows and basic balance surpluses to
accumulate reserves. However; after the Asian crisis, capital flight gained
importance. More basic balance surpluses were needed to cover this flight. As NI
deficits continued, surplus in BENI has increased after 1997.
4.2.10. Thailand
Thailand is one of the second-generation Asian tigers and at the end of 1980s it had
growth rates around 10%. Financial liberalization first targeted inflows. Restrictions
on inflows of portfolio investment were eliminated between 1985 – 87.
82
Figure 20. Ratio of BOP components to external debt stock: Thailand
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
0.4
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
NIBENIRKODF
Asian crisis
liberalization
Source: author’s calculations, IFS
Figure 20 shows results of Exercise 1 for Thailand. Ease of restrictions on capital
outflows began in 1991. The establishment of Bangkok International Banking
Facility, which originally aimed strengthening of domestic financial sector by
fostering its linkages with other Southeast Asian financial sectors, induced more
external borrowing and eased outflow of resident capital. Authorities responded to
huge rises in debt flows by imposing reserve requirements on inflows in 1995. These
controls, however, have been ineffective. The debt flows fell sharply in 1997 when
Asian crisis began in Thailand. Authorities responded the crisis with Malaysian style
controls on outflows. Lending to non-residents in local currency was restricted. Yet,
controls are lifted a year later.
83
Reflecting the low level of debt burden of Thailand, NI deficit was small compared
to debt flows throughout the period. However, like Korean case, large current
account deficits, given low values of foreign direct investment culminated in BENI
deficits. Consequently, large values of basic balance deficits constituted a major
financing need in Thailand. Meanwhile, Thai authorities accumulated significant
amount of reserves after liberalization. This reserve accumulation has been
consistent until the attacks on reserves in the Asian crisis. Capital outflows were
negative before 1992, when they were restricted. With the establishment of Bangkok
International Banking Facility, capital flight has become another item that required
financing with debt flows.
Table 15 presents share of each balance of payments component in finance through
debt flows. Values are given for three periods. The first column covers all the debt
flows pour period. The second column covers the period after the opening of
Bangkok International Banking Facility when the capital flight gained importance.
The last column cover the period 1990 – 96 and given to provide comparison
opportunity with the analyses of the other Asian countries. Large basic balance
deficits constitute the most important financing item after liberalization. More than a
half of debt flows were used to finance basic balance deficits. The second important
item has been reserve accumulation. 38% of debt flows were used to accumulate
reserves between 1987 – 1996. The importance of capital outflows increase after
1993, as explained above.
84
Table 15. Usage of Debt Flows in Thailand (1987 – 1996)
1990 - 1996 1993 – 1996 1987 – 1996 Basic balance deficit 60% 59% 57% Outflow of resident capital 7% 13% 5% Reserve accumulation 32% 27% 38%
Source: author’s calculations, IFS
After 1997, debt flows give consistently negative balances. With devaluation, BENI
values adjusted to give surpluses. Capital flight continued after the crisis. Reserve
accumulation has been much lower than the pre-crisis period. In the post-crisis
period, large basic balance surpluses were used to finance capital flight and debt
repayments (negative values in debt flows).
4.2.11. Indonesia
Indonesia has been one of the second-generation Asian tigers. The growth rates have
been high in 1980s. Financial liberalization attempts began in late 1980s and
completed to a large extent by 1989.
Figure 21 demonstrates results of Exercise 1 for Indonesia. Debt flows have been
consistently positive until the Asian crisis of 1997. However, level of debt flows as a
proportion of external debt stock is below 10%, which is lower than the value of
many emerging markets. NI balance was negative and positive BENI values was not
enough to offset them resulting in a basic balance deficit throughout the period.
Figure 21. Ratio of BOP components to external debt stock: Indonesia
85
-0.15
-0.1
-0.05
0
0.05
0.1
0.15
0.2
0.25
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
NIBENIRKODF
Asian crisis
liberalization
Source: author’s calculations, IFS
Nearly half of the debt flows attracted in the period of 1990 - 96, as demonstrated in
Table 16, were used to finance this deficit. Capital flight and reserve accumulation
also had significant portions in debt financing, 19% for the former and 32% for the
latter. Sharp fall in debt flows in 1997 and negative values afterwards, resulted in a
rise in BENI surpluses. Consequently, basic balance adjusted to positive values.
After 1997, most of the capital outflows were reversed, while reserve accumulation
continued.
Table 16. Usage of Debt Flows in Indonesia (1990 – 1996)
Basic balance deficit 48%Outflow of resident capital 19%Reserve accumulation 32%
Source: author’s calculations, IFS
86
4.2.12. Philippines
Philippines began its capital account liberalization attempts in early 1980s, but
significant liberalization was accomplished between 1992 – 95.
Figure 22. Ratio of BOP components to external debt stock: Phillipines
-0.30
-0.20
-0.10
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
R
KO
NI
BENI
DF
liberalization
crisis
Source: author’s calculations, IFS
Figure 22 discloses results of Exercise 1 for Philippines. Surge in debt flows
accompanied this liberalization. As a different case from other emerging market
countries, basic balance deficits of Philippines stemmed from negative BENI values,
not NI deficits. Capital flight was small at the early years of liberalization; however
it has increased after 1995. Reserve accumulation has also been a financing need
throughout the liberalization period. The effect of Asian crisis on Philippines has
been short-lived. Debt flows recovered after a temporary reduction in 1998. Yet,
87
after 1998, capital flight surged and resident capital outflows overweighed debt
flows. Philippines, therefore, gave basic balance surplus to finance this capital flight.
Table 17. Correlation of DF with other components in Philippines
1978 - 1990 1991 – 1997 BB 0.59 0.15 BB(-1) 0.58 0.10 BB(+1) 0.52 0.23 KO 0.53 0.64 R 0.33 0.66 Source: author’s calculations, IFS
Table 17 illustrates usage of debt flows between 1989 and 1997. The figures in the
second column are presented to provide comparison opportunity with other Asian
emerging market analysis. Basic balance deficit and capital outflows have nearly the
same weight around 40%. The remaining portion of debt flows financed reserve
accumulation.
Table 18. Usage of Debt Flows in Philippines (1989 – 1996)
1989 - 19971990 –
1996Basic balance deficit 39% 33%Outflow of resident capital 37% 29%Reserve accumulation 22% 33%
Source: author’s calculations, IFS
Table 18 shows correlation values between debt flows and other components before
and after capital account liberalization and surge in debt flows. The correlation
between debt flows and basic balance falls from 0.59 to 0.15. On the other hand,
88
correlation values between debt flows and capital outflows / reserve accumulation
increases. These values suggest a break-up in the relationship between basic balance
finance and debt flows after liberalization.
4.2.13. Hungary
Hungary has rapidly liberalized its current account and foreign direct investments
after the fall of iron curtain in 1989. Nevertheless, liberalization of portfolio and
other capital flows has been more gradual. First attempts date back to early 1990s,
but the significant liberalization came with OECD membership in 1996. Even after
OECD membership, some transitional clauses applied for some OECD codes on
capital flows and liberalization completed in 2001.
Results of Exercise 1 for Hungary are shown on Figure 23. Opening up of
Hungarian market has resulted in a surge in debt flows, in accordance with other
transition countries, in early 1990s. This surge ended in 1996. The other heavy debt
flows period began in 1998 and lasted until 2002. In this period, debt flows were
attracted not only by larger liberalization measures, but also relative confidence
provided by the EU accession process, which began in 1996.
The income account of Hungary has consistently gave negative balance, as it had a
large debt burden and it is traditionally an attractive point for foreign direct
investment. After 1997, with EU accession process, attraction of more FDI resulted
in surpluses in BENI, partially covering deficit in NI. In both periods of surge in
89
debt flows, Hungary accumulated large amounts of reserves. Capital flight has been
positive between 1996 – 1999. As demonstrated in Table 19, in the debt flows surge
after 1996, 53% of debt flows financed reserve accumulation, 15% financed capital
outflows and 32% financed basic balance deficits. If the whole experience after 1993
is considered, the portion used for reserve finance rises to 82%. After 2001, current
account deficit widened making BENI negative, although FDI inflows continued.
With NI also in deficit, basic balance gave large negative values crowding out the
finance provided to other components by debt flows.
Figure 23. Ratio of BOP components to external debt stock: Hungary
-0.25
-0.2
-0.15
-0.1
-0.05
0
0.05
0.1
0.15
0.2
0.25
0.3
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
NIBENIRKODF
liberalization
Source: author’s calculations, IFS
90
Table 19. Usage of Debt Flows in Hungary (1992 - 1998)
1993 - 2001 1998 - 2001Basic balance deficit 9% 32%Outflow of resident capital 9% 15%Reserve accumulation 82% 53%
4.2.14. Czech Republic
Czech Republic adopted a rapid liberalization strategy for its capital account, after
communism ended and the former Czechoslovakia broke up into two nations.
Restrictions on inflows were eliminated between 1993 – 1995. Some restrictions on
outflows remained however, and those were mostly eased in 2001.
Figure 24. Ratio of BOP components to external debt stock: Czech Republic
-0.60
-0.40
-0.20
0.00
0.20
0.40
0.60
0.80
1.00
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003
NIBENIRKODF
liberalization
Source: author’s calculations, IFS
91
Figure 24 demonstrates results of Exercise 1 for Czech Republic. Debt flows poured
with liberalization after 1993. Czech Republic used most of the debt flows to
accumulate reserves. As the investor confidence in property rights have not yet
established early liberalization period witnessed flight of resident capital, although
they were legally restricted. A slowdown in capital inflows resulted in a currency
crisis in 1997. Debt flows have been reduced after the crisis and remained stable in
low levels.
A unique feature of Czech Republic has been the low level of burden of past debt.
Consequently, unlike many emerging market countries, the deficit in NI has been
very low. Moreover, this deficit has been surpassed by positive BENI balances. The
BENI surpluses rose significantly after 1998, when EU accession process began and
Czech Republic has become one the most FDI-attracting countries as a proportion of
the size of its economy in the world. Consequently, throughout the liberalization
period the basic balance has a positive balance. Capital flight also declined as
domestic investors restored confidence to the country. A significant portion of the
basic balance surplus, thus, used to accumulate reserves.
4.2.15. Poland
Poland’s capital account liberalization has been parallel to the developments in
country’s relationships with the EU, IMF and OECD. Poland adopted the Article
92
VIII of IMF agreement in 1995. It has been a member of the OECD in 1996. With
these developments, most restrictions on capital movements have been removed.
Poland’s capital account regulations have been conformable with EU when it
became a member in 2004.
Figure 25 shows results of Exercise 1 for Poland. The surge in debt flows to Poland
take place after 1996. In the first tow years, nearly all debt flows finance reserve
accumulation. Except 1999, the BENI always gives positive balances, in contrast
with the deficit values of NI. After 1999, capital outflows constitute and important
financing need while the role of reserve accumulation becomes small.
Figure 25. Ratio of BOP components to external debt stock: Poland
-0,4
-0,3
-0,2
-0,1
0
0,1
0,2
0,3
0,4
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
NIBENIRKODF
liberalization
Source: author’s calculations, IFS
93
Table 20. Usage of Debt Flows in Poland (1997 – 2003)
Basic balance deficit 29%Outflow of resident capital 35%Reserve accumulation 36%
Source: author’s calculations, IFS
In Poland, the surge in debt flows mostly financed capital outflows and reserve
accumulation. The figures in Table 20 demonstrate this phenomenon. The share of
capital flows and reserve accumulation within finance created by debt flows are
almost the same at 35%. The remaining portion of debt flows financed basic balance
deficits.
4.3. Concluding Remarks
Some observations on country analyses presented in Section 4.2 are worth noting.
The analyses suggest that, when the behaviors of the new structural components are
concerned, Asian and East European countries are more homogenous compared to
Latin America. Comparing the end use of debt flows; roughly speaking, in Asian
countries basic balance finance and in East European countries (after 1995) reserve
accumulation has been important. Latin American countries provide different
pictures. Capital outflows dominate other items in Argentina, Chile and Brazil;
whereas basic balance finance is important in Mexico and Columbia. Meanwhile,
outflows of resident capital in aggregate give negative balance in Peru, i.e., residents
imported their capital back to country. An inter-regional comparison, on the other
94
hand, shows that the dominant item in the end use of debt flows had been outflows
of resident capital in Latin America, basic balance deficits in Asia and reserve
accumulation in East Europe.
Capital outflows, when aggregated for all the period after liberalization, gave
positive balance in all countries, except Peru. Liberalizing capital account rapidly or
gradually does not create much difference in this sense. A simple comparison of
rapid liberalizer Argentina with gradual liberalizer Brazil; or rapid liberalizer
Thailand with gradual liberalizer Korea suggests that the outflow of resident capital
is due to domestic imbalances or inappropriate management of capital account rather
than the pace of liberalization. Even in Venezuela, which had sporadic repression
periods in its capital account, a high amount of capital flight was recorded.
Moreover, a careful examination of patterns of capital outflows reveal certain
triggering events play an important role in raising the amount of capital outflows.
These events can be one-shot liberalization of outflows (like Argentine liberalization
in 1991) or some other regulatory decisions. The examples of the latter are
allowance for pension funds to invest abroad in Chile (1998) and establishment of
Bangkok International Banking Facility in Thailand (1993). Capital outflows
increase sharply with these triggering factors and then they “fade out.”
In each country analyzed, with the exception of Chile, we observe sudden stops in
capital flows. In these crisis years, capital flows fell to negative values, i.e., the
emerging economy concerned became a net creditor. The adjustment of other
95
components to falls in debt flows is worth analyzing. In almost all instances, except
Indonesia (1997) and Hungary (2002), no reversal in capital flight occurred. Put it in
another way, residents did not take their capital back to country to pay back their
external debt. Instead, debt was paid either from the reserves, or by adjusting BENI
to surplus values.
Noting the exceptional circumstances elaborated in Section 4.231, the reviews of
country experiences offer a general picture after capital account liberalization:
Immediately after liberalization, debt flows increase dramatically. Then, first capital
outflows and sequentially reserve accumulation increase. A deficit in the NI balance
accompanies this process. When debt flows stop, as NI is typically exogenous and
resident capital rarely returns, after an initial loss in reserves, BENI starts to give
surplus balances.
After liberalization, significant portion of debt flows has been used to finance capital
outflows and reserve accumulation. A structural income account deficit constituted
most of the remaining part of the basic balance deficit financed. As a result, the
emerging market economies, on average, did not run deficits in the remaining parts
of the basic balance, i.e., they have not been able to borrow to buy investment goods
that may be used to enhance their productive capacities.
31 Some notable exceptional circumstances are absence of capital outflows in Peru, very low NI deficits and atypical BENI deficits in Korea, large BENI surpluses in Venezuela after 1995 due to oil revenues.
96
CHAPTER V
A CASE STUDY: TURKEY
In this chapter I analyze the Turkish experience with financial liberalization using
the framework provided in Chapter 3. Turkey provides a good example where
capital account liberalization did not bring high growth rates: The average annual
growth rate of real GDP has been 3.59% between 1990 - 2003 compared to 4.39% of
1970 – 89 pre-liberalization period. In the first section, I repeat the exercises
undertaken in Chapter 4. In the second section, I provide a political economy
perspective to analyze the relationship between components of balance of payments
and establish a causal link between them using vector auto-regression (VAR)
analysis. In the third section, I compare Turkey with the other countries analyzed in
Chapter 4 and I provide some conclusions from Turkish experience.
5.1. A Structural Analysis of Turkish Balance of Payments Accounts
The results of the Exercise 1 are depicted in Figure 26 where contributions of each
of our new components can be traced for the 1981 – 2003 period. Obviously,
Turkey has been a net borrower since 1981 except the crisis years of 1994 and
97
2001, i.e., the debt flows gave positive balance. Sudden stops in capital inflows can
easily be observed in the figure as sharp falls from 1993 and 2000’s positive values
to the following year’s negative values.
Figure 26. Ratio of BOP components to external debt stock: Turkey
-0,30
-0,20
-0,10
0,00
0,10
0,20
0,30
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
NIBENIRKODF
I II III
liberalization crisis
Source: author’s calculations, IFS
Table 21. Usage of debt flows in Turkey (1981 – 2000)
1981 - 1989 1990 - 1993 1995 - 2000 Basic balance deficit 64% 32% 34% Outflow of resident capital 25% 61% 30% Reserve accumulation 11% 07% 36%
Source: author’s calculations, IFS
There are three patterns in usage of debt flows in 1981 – 2001 period. The first
phase is observed until the financial liberalization of 1989. Before liberalization,
basic balance deficits were mostly financed through debt flows. Financial
98
liberalization represents the beginning of the second phase. Liberalization brought a
flight of resident capital. The results of the Exercise 2 presented in Table 26 show
the change in patterns more clearly. The portion of debt flows used to finance
capital outflows rose from 25% to 65% from first phase to the second, whereas the
share of basic balance deficit fell from 64% to 32%.
1994 marks another shift in the pattern. Being trapped in the financial crisis with a
low level of reserves, macroeconomic authorities then gave the utmost importance
to having a high reserve level and debt flows were used mostly for financing reserve
accumulation. In the third phase, the portion of debt flows used to finance reserves
accumulation rise to 36% from 7%, while portion used finance of capital outflows
fall to 34%. In 1994 and 2001, debt flows became negative, i.e., Turkey became a
net exporter of capital. In both crises years, basic balance surpluses played a
significant role in financing the flight of non-resident capital. In 1994 crisis, a
reversal in outflow of resident capital financed the remaining portion of non-
resident capital flight; whereas in 2001 such a reversal did no occur, instead of that,
a reduction in reserves accompanied outflow of non-resident capital.
Apparently, basic balance deficit has been mostly due to the deficit in income
account. Without the deficit in income account, the BENI component does not give
notable deficits except 1993 and 2000, the years of consumption booms preceding
financial crisis. Within BENI the most important item is current account balance
excluding income account since Turkey has been a recipient of very little FDI. The
99
ratio of FDI to GDP never exceeds 1%, except 200132. Accordingly, 96% of the
deficit in income account has been due to interest paid to foreign debt stock both in
1984 - 2001 and 1990 - 2001 periods. Financial liberalization did not help non-
debt-generating capital inflows to increase, instead large interest payments to the
existing stock continued to be a burden and needed to be financed after
liberalization.
The above analysis reveals that the traditional link between basic balance deficit
finance and debt flows lost its importance after capital account liberalization. After
Turkish financial liberalization of 1989, rather than financing basic balance deficits,
debt flows were mostly used to finance outflow of resident capital, reserve
accumulation and interest paid to existing short-term debt stock. Obviously, none of
these three components financed by debt flows has a positive effect on the
productive capacity of the country. Consequently, financial liberalization did not
fulfill the neo-liberal premise of delivering a high growth rate. In the next section,
we will analyze the Turkey's experience in detail.
32 The only large figure in 2001 is an exception due to transfer of a new mobile phone license to a foreign operator and acquisition of a Demirbank, a resident bank, by HSBC.
100
101
5.2. Turkey's Experience with an Open Capital AccountTP
33PT
When Özal government took the liberalization decision in 1989, it was most
probably an attempt to curtail the macroeconomic imbalances reappearing after the
export-led growth boom of 1980s (Akyüz and Boratav, 2002; Alper and Öniş, 2001).
From 1986 to 1989, public sector borrowing requirement to GDP ratio (PSBR) had
already risen from 3.7% to 5.3% and inflation increased from 36% to 49%. Yet, the
surge in capital inflows after liberalization aggravated the imbalances: Public sector
borrowing requirement (PSBR) had been consistently on the rise and the
government had opted for the easy debt-financing way to sustain these high levels
by the help of abundant capital inflows. The debt-financing was in indirect form:
government mostly borrowed in domestic markets, while domestic financial agents
borrowed from abroad, increasing the external debt stock eventually. Indirect
financing did not make a difference about solvency: Had the government become
insolvent, its domestic lenders would be insolvent too. In addition, as Öniş and
Aysan (2000) put forward, government had been reluctant to take necessary
measures to increase its revenues (such as tax reform) as it was presumed that the
debt-finance would be never-ending. By 1993, PSBR was 12% and inflation rate
was 65%.
TP
33PT See Cizre-Sakallioglu and Yeldan (2000), Boratav and Yeldan (2002) and Alper and Onis (2001) for
detailed analysis about the post-liberalization Turkish economy. Güven (2001) provides a review of the reasons and effects of capital movements in the Turkish context.
High inflation rates have been another factor accompanying public deficits. The
inflation in Turkey is mostly due to expectations based on past levels. Moreover,
high rates of inflation raise the volatility of inflation itself. During 1990s, the
inflation rate averaged 75% while its standard deviation has been 13%. Volatility in
inflation, naturally, caused volatility in the real return of investments in TL
denominated assets. In spite of the fact that TL denominated assets have real returns
averaging 30%; because of the volatility, Turkish residents fled from TL and
consistently hedged to foreign currencies by investing in foreign currency
denominated assets which have been perfect substitutes to TL denominated assets
thanks to liberalization. This flight took two forms: First, they invested abroad to
avoid default risk of government and domestic banks, which resulted in outflows of
resident capital. Second, they invested in foreign currency assets in Turkey to avoid
the exchange rate risk of staying in TL denominated assets. The currency
substitution stemming from the second form of hedging can be observed from the
M2Y / M2 ratio. This ratio increased from 143% in 1992 to 192% in 1995 and to
225% in 2001. As a result, foreign currency deposits dominated TL deposits, raising
their share in total deposits from 17% in 1987 to 49% in 1995. In 2001, more than
half of the deposits in Turkish banks were in foreign currency terms. Both the
capital flight from residents and increase in foreign currency deposits are reflected in
the rise in "capital outflows" aggregate after liberalization. The capital flowed into
the country was directed mostly to finance domestic agents' hedging.
102
I run a VAR to clarify the relationship between debt flows and capital outflows.
Using monthly data for 1992 January - 1998 October period, I run two VARs, one
with a lag length of 4 and one with 12. The former lag length is also the one offered
by AIC and SIC. The results can be interpreted as causality relations in the short-
and medium- run, respectively.
Figure 27. Accumulated impulse response functions (DF and KO, short lag length)
-200
0
200
400
600
800
1000
5 10 15 20
Accumulated Response of KO to KO
-200
0
200
400
600
800
1000
5 10 15 20
Accumulated Response of KO to DF
400
800
1200
1600
5 10 15 20
Accumulated Response of DF to KO
400
800
1200
1600
5 10 15 20
Accumulated Response of DF to DF
Accumulated Response to Cholesky One S.D. Innovations
Source: author’s calculations, Central Bank of Turkey; Lag length: 4 months
103
According to the VAR results, in the short-run there exists two-way (Granger)
causality between debt flows and capital outflows. In the long run, on the other
hand, only debt flows (Granger) causes capital outflows. International investors
make their decisions according to the return rates in industrial economies and
financial situation in other emerging market countries. Because of the contagion and
herding effects, it is natural for debt flows to be exogenous in medium- and long-run.
Figure 27 and 28 demonstrate the impulse response functions. As expected,
accumulated effect of debt flows on capital outflows is positive for both lag lengths,
that is, debt borrowed from non-residents raises the outflow of resident capital34.
Another by-product of capital account liberalization has been a consistent overvaluation
of TL. Apparently, after 1989 the real exchange rate of TL against major currencies
moved to a lower plateau and except the depreciations in the crises years, stayed there.
TL had an average real appreciation of 25% against US dollar in 1990s. The
overvaluation has perverse effects on exports and growth. In addition, overvaluation
had a significantly positive effect on imports. Moreover, as debt stock was augmented,
the interest paid for it rose too. Through these two channels, debt flows once used to
finance basic balance deficits before liberalization, became a stimulating factor for
those deficits.
34 The causality relationship is robust to exclusion of “Net Errors and Omissions” and / or "currency holdings of banks" from capital outflows aggregate.
104
Figure 28. Accumulated impulse response functions (DF and KO, long lag length)
-200
0
200
400
600
800
1000
1200
5 10 15 20
Accumulated Response of KO to KO
-200
0
200
400
600
800
1000
1200
5 10 15 20
Accumulated Response of KO to DF
0
400
800
1200
1600
2000
5 10 15 20
Accumulated Response of DF to KO
0
400
800
1200
1600
2000
5 10 15 20
Accumulated Response of DF to DF
Accumulated Response to Cholesky One S.D. Innovations
Source: author’s calculations, Central Bank of Turkey; lag length: 12 months
In order to analyze the relationship between debt flows and basic balance deficits, I ran a
VAR equation with 12 lags and using monthly data for 1992 - 2003 period. The lag
value of 12 is offered by various criteria, but the results are robust for different lag
values. According to the VAR results, debt flows (Granger) cause basic balance
105
deficits35. As observed from the impulse response functions presented in Figure 29, the
deficit effect is strong in the last 10 months.
Figure 29. Accumulated impulse response functions (DF and BB)
0
200
400
600
800
1000
1200
2 4 6 8 10 12 14 16 18 20
Accumulated Response of BB to BB
0
200
400
600
800
1000
1200
2 4 6 8 10 12 14 16 18 20
Accumulated Response of BB to DF
0
400
800
1200
1600
2000
2 4 6 8 10 12 14 16 18 20
Accumulated Response of DF to BB
0
400
800
1200
1600
2000
2 4 6 8 10 12 14 16 18 20
Accumulated Response of DF to DF
Accumulated Response to Cholesky One S.D. Innovations
Source: author’s calculations, Central Bank of Turkey; lag length: 4
Although debt-financing was an easy way to finance public deficits, it was soon
understood that it was a never-ending process. As continuity of capital inflows
35 VAR results are robust for different lag length specifications and for seasonally adjusted values of the basic balance deficit.
106
depended on high arbitrage levels for non-resident investors, while borrowing in high
interest rates, government repressed depreciation of TL. As a result, arbitrage levels
have generally been higher than 20% and rose occasionally to rates over 60% (see
Figure 30). However, when investor expectations were reversed, it had been
impossible to access to international financial capital even at high arbitrage levels.
Boratav (2001) points out that the swing in debt inflows reached 25.6 billion US
dollars in the major debt flow reversal of 2001 crisis.
Meanwhile, high arbitrage levels did not result in long-term borrowing. For the 1990
- 1999 period, short-term flows comprise 74% of the total debt flows36. The
existence of a short-term debt stock reduced investor confidence and raised interest
burden more. The short-term debt structure resulted in a Ponzi-type debt financing
and refusal of roll over by investors has been a triggering effect for financial crises.
In the second half of 1993, to prevent a further rise in the cost of servicing the
domestic debt, the government cancelled various domestic debt auctions and
accepted a small percentage of short maturity offers. Reversing the investor
expectations, this attempt resulted in a run on foreign currency and consequently
the 1994 financial crisis (Özatay, 1999). After 1994, the government did not have
any attempt to solve the short-termism in debt financing. Rather, continuing in
short-term borrowing, the government implemented a de facto managed float regime
repressing depreciation to attain high arbitrage levels.
36 After the implementation of IMF program, most of the short term debt was converted to IMF credits and this ratio falls to 48% for 1990 - 2003 period.
107
Figure 30. Rate of financial arbitrage in Turkey (%)
-60
-40
-20
0
20
40
60
80
100
120
Jan.92 Sep.92 May.93 Jan.94 Sep.94 May.95 Jan.96 Sep.96 May.97 Jan.98 Sep.98 May.99 Jan.00 Sep.00 May.01
Source: author’s calculations, Central Bank of Turkey Rate of financial arbitrage = (1 + nominal interest rate) / (1 + depreciation of TL)
Caught to the 1994 crisis with a low level of reserves, the government began to use
reserve accumulation as a shield to avoid depreciation and against another crisis. As
a result, the costs of reserve accumulation, as elaborated in Section 2.3.2, fell on the
country.
The combination of the circumstances explained above resulted in a vicious
cycle of short-term debt, high interest rates, overvalued TL, and a Ponzi-type
debt financing. Persistent outflow of resident capital, currency substitution and
costly reserve accumulation accompanied this vicious cycle. By 1999, it
became evident that public debt was at unsustainable levels. Meanwhile, PSBR
108
already reached 15% and fragility indicators relating to short-term debt were at
all-time high levels. A last attempt to control the debt stock by the IMF-led
stabilization program of 1999 resulted in the financial crisis in February 200137.
5.3. Concluding Remarks
Turkish capital account liberalization experience witnessed two financial crises. The
period before the 1994 crisis resembles a Latin American style liberalization style,
where large outflows of resident capital came after the liberalization decision. In this
sense, 1989 liberalization can be compared with the contemporaneous liberalization
of Argentina. Nevertheless, in 1994, a reversal in capital flight occurred, i.e.,
residents took their capital back to use it in paying back debt. This reversal in capital
outflows is an exception in responses of emerging market countries to sudden stops
in capital flows. On the other hand, the period after 1994 is much like Asian patterns
of behavior of balance of payment components. A large reserve accumulation took
place in this period. In the 2001 reversal in debt flows, much of these reserves were
lost and when the exchange rate peg was given up, BENI adjusted by giving surplus.
According to the Turkish capital account liberalization experience, premature
liberalization resulted in a flight of resident capital. Consequently, after capital
37 There were two major elements in the stabilization program: a currency peg to a basket of euro and dollar and a quasi-currency board where expansion of monetary base was limited to expansion of net foreign assets. See Yeldan (2002), Akyüz and Boratav (2002) and Sak and Ozatay (2003) a variety of analysis of financial crisis of 2001.
109
account liberalization, the traditional link between basic balance deficit finance and
debt-generating inflows of non-resident capital lost its importance. Debt flows
became a cause of outflow of non-resident capital. My VAR analysis shows that
debt flows (Granger) cause outflows of resident capital. In addition, as overvaluation
of TL accompanied debt flows, they became a source of basic balance deficits,
rather than a means to finance it. According to my VAR results, demonstrates that
debt flows (Granger) cause basic balance deficits. After liberalization, instead of
basic balance deficits, debt flows were mostly used to finance reserve accumulation
and interest paid to existing short-term debt stock. Used in an unproductive way, it
had not been possible for the debt flows to bring more economic growth.
110
CHAPTER VI
CONCLUSION
In this thesis, I argue that, in emerging market countries, after capital account
liberalization the traditional link between the debt flows and finance of basic balance
deficits lost its importance. I decompose the balance-of-payments and construct new
components that make it possible to track the new financing patterns. I analyze
emerging markets in three groups: Latin America, Asia and East Europe. Then, I
make individual analysis for 15 countries and a special case study for Turkey.
Each region and country has its own circumstances regarding international financial
liberalization. Nevertheless, in the framework of the analysis presented in this thesis,
it is possible to argue that premature liberalization and inappropriate
macroeconomics policies led to a new financing pattern where debt flows mostly
covered mostly financed outflows of resident capital and reserve accumulation
instead of the structural financing needs of the emerging markets. In addition, many
emerging market countries actually had surpluses in their basic balances if interest
and profit transfers to abroad are excluded. As a result of this new financing pattern,
111
the positive effect of capital account liberalization in closing the investment – saving
gap in emerging markets has been limited.
The framework provided in this thesis provides a new perspective in explaining the
lack of a robust link between capital account liberalization and growth. The capital
inflows after liberalization, however large, have financed a large capital flight and a
considerable unproductive saving as reserve accumulation. Therefore, capital
inflows merely augmented national savings used in productive ways. This new
dimension may complement the existing explanations in the liberalization and
growth literature.
112
113
SELECT BIBLIOGRAPHY
Acemoğlu, Daron, James A. Robinson and Simon Johnson. 2001. “The Colonial Origins of Comparative Development: An Empirical Investigation” American Economic Review. 91: 1369-1401.
Aizenman, Joshua, Jaewoo Lee. 2005. “International Reserves: Precautionary versus Mercantilist Views, Theory and Evidence,” Unpublished manuscript. University of California at Santa Cruz.
Akyüz, Yılmaz and Korkut Boratav. 2002. “The Making of the Turkish Financial Crisis.” UNCTAD Discussion Papers 158. Geneva: United Nations Conference on Trade and Development.
Alfaro, Laura, Sebnem Kalemli-Ozcan and Vadym Volosovych. 2005. “Capital Flows in a Globalized World: The Role of Policies and Institutions.” Paper presented at NBER Conference on Capital Controls and Capital Flows in Emerging Economies: Policies, Practices and Consequences. December 16-18, 2004.
Alper, Emre and Ziya Öniş. 2001. “Financial Globalization, the Democratic Deficit and Recurrent Crises in Emerging Markets: The Turkish Experience in the Aftermath of Capital Account Liberalization. ISS/EC Working Paper No. 01-14. İstanbul: Boğaziçi University.
Ariyoshi, Akira, Karl Habermeier, Bernard Laurens, İnci Ötker-Robe, Jorge Ivan Canales-Kriljenko and Andei Kirilenko. 2000. Capital Controls: Country Experience with Their Use and Liberalization. IMF Occasional Paper 190. Washington, DC: International Monetary Fund.
Arteta, Carlos, Barry Eichengreen, and Charles Wyplosz. 2001. “When Capital Account Liberalization Help More Than It Hurts?” NBER Working Paper 8414. Cambridge, MA: National Bureau of Economic Research.
Bekaert, Geert, Campbell R. Harvey, Christian Lundblad. 2001. “Does Financial Liberalization Spur Growth?” NBER Working Paper 8245. Cambridge, MA: National Bureau of Economic Research.
114
Boratav, Korkut and Erinç Yeldan. 2002. "Turkey, 1980 - 2000: Financial Liberalization, Macroeconomic (In)stability and Patterns of Accumulation" Unpublished manuscript. CEPA and New School for Social Research.
Boratav, Korkut. 2001. "2000 / 2001 Krizinde Sermaye Hareketleri" İşletme, İktisat ve Finans, 186: 7-17. Eylül.
Boratav, Korkut. 2003. "Yabanci Sermaye Girişlerinin Ayrıştırılması ve Sıcak Para: Tanımlar, Yöntemler, Bazı Bulgular" in Ahmet Haşim Kose, Fikret Şenses and Erinç Yeldan, eds., İktisadi Kalkınma, Kriz ve İstikrar, İstanbul: Iletişim Publications.
Bordo, Michael, Bary Eichengreen and Jong Woo Kim. 1998. Was there really an earlier period of international financial integration comparable to today? NBER Working Paper 6738. Cambridge, MA: National Bureau of Economic Research.
Calvo, Guillermo. 1998. “Capital Flows and Capital-Market Crises: The Simple Economics of Sudden Stops,” Journal of Applied Economics, 1 (1): 35-54, November.
Cizre-Sakallıoğlu, Ümit, Erinç Yeldan. 2000. "Politics, Society and Financial Liberalization: Turkey in the 1990s," Development and Change, 31(2): 481-508.
Claessens, Stijn and David Naude. 1993. “Recent Estimates of Capital Flight.” World Bank Policy Research Working Papers 1186. September. Washington, DC: World Bank.
Cornfold A. and J. Brandon. 1999. “The WTO Agreement of Financial Services: Problems of Financial Globalization in Practice” in UNCTAD, International Monetary and Financial Issues for 1990s, vol X, New York and Geneva.
Cowan, Kevin and José De Gregorio. 2005. “International Borrowing, Capital Controls and the Exchange Rate: Lessons From Chile.” Paper presented at NBER Conference on Capital Controls and Capital Flows in Emerging Economies: Policies, Practices and Consequences. December 16-18, 2004.
De Gregorio, Jose Sebastian Edwards, Rodrigo O. Valdes. 2000. “Controls on Capital Inflows: Do they Work?” NBER Working Paper 7645. Cambridge, MA: National Bureau of Economic Research.
De Souza, Lfflcio Vinhas. 2004. “Financial Liberalization and Business Cycles: The Experience of Countries in the Baltics and Central Eastern Europe.” Deutsche Bundesbank Discussion Paper series 23/2004. Frankfurt: Deutsche Bundesbank
Dominguez, Kathryn M.E. and Linda L. Tesar. 2005. Paper presented at NBER Conference on Capital Controls and Capital Flows in Emerging Economies: Policies, Practices and Consequences. December 16-18, 2004.
Dooley, Michael. 1988. “Capital Flight: A Response to Differences in Financial Risks,” IMF Staff Papers. 35: 422 – 36. September.
Dooley, Michael and Kenneth. M. Kletzer. 1994. “Capital Flight, External Debt, and Domestic Policies” Economic Review of Federal Reserve Bank of San Francisco 3: 29 – 37.
Edison, Hali J., Michael Klein, Luca Ricci, and Torsten Slok. 2004. “Capital Account Liberalization and Economic Performance: Survey and Synthesis,” IMF Staff Papers. 51(2): 220 – 256.
Edison, Hali. J., Ross Levine, Luca Ricci, Torsten Slok. 2002. “International Financial Integration and Economic Growth.” NBER Working Paper 9164. September. Cambridge, MA: National Bureau of Economic Research.
Edwards, Sebastian. 1999. “How Effective are Capital Controls?” NBER Working Paper 7413. Cambridge, MA: National Bureau of Economic Research.
Edwards, Sebastian. 2001. “Capital Mobility and Economic Performance: Are Emerging Economies Different?” NBER Working Paper 8076. Cambridge, MA: National Bureau of Economic Research.
Eichengreen, Barry J. 2003. “Capital Account Liberalization: What Do Cross-Country Studies Tell Us?” in Barry Eichengreen, ed., Capital Flows and Crisis. Cambridge, MA. MIT Press, 71 – 98.
Eichengreen, Barry J. And Michael Mussa. 1998. Capital Account Liberalization: Theoretical and Practical Aspects. IMF Occasional Paper 172. Washington, DC: International Monetary Fund.
Eichengreen, Barry, and David Leblang. 2002. “Capital Account Liberalization and Growth: Was Mr. Mahathir Right?” NBER Working Paper 9427. Cambridge, MA: National Bureau of Economic Research.
Feldstein, Martin S. 1999. “Self-Help Guide for Emerging Markets” Foreign Affairs, 78: 93 – 100.
Ghosh, Jayati. 2005. “The Economic and Social Effects of Financial Liberalization. A Primer for Developing Countries.” Unpublished manuscript. Available: www.networkideas.com
115
116
Goldfajn, Ilan and André Minella. 2005. “Capital flows and controls in Brazil: What have we learned?” Paper presented at NBER Conference on Capital Controls and Capital Flows in Emerging Economies: Policies, Practices and Consequences. December 16-18, 2004.
Goldsmith, Raymond W. 1969. Financial Structure and Development. New Haven, CN: Yale University Press.
Gourinchas, Pierre-Olivier, and Olivier Jeanne. 2004. “The Elusive Gains from International Financial Integration” IMF Working Paper No. 04/74. Washington, DC: International Monetary Fund.
Güven, Semih. 2001. "Sermaye Hareketlerinin Nedenleri, Etkileri ve Türkiye Örneği" İşletme, İktisat ve Finans, 185: 79 - 83. Ağustos.
Hall, Robert, and Charles Jones. 1999. “Why Do Some Countries Produce So Much More Output per Worker Than Others?”. Quarterly Journal of Economics. 114: 83-116. February.
IMF. 1995. Balance of Payments Manual. Washington, DC.
IMF. 1996. Balance of Payments Textbook. Washington, DC.
IMF. 2005. Report on the Evaluation of the IMF's Approach to Capital Account Liberalization. Washington, DC.
Ishii, Shogo and Karl Habermeier. 2002. Capital Account Liberalization and Financial Sector Stability. IMF Occasional Paper 211. Washington, DC: International Monetary Fund.
Johnston, R. Barry, Salim M. Darbar, and Claudia Echeverria. 1997. “Sequencing Capital Account Liberalization: Lessons from Experiences in Chile, Indonesia, Korea and Thailand.” IMF Working Paper 97/157. Washington, DC: International Monetary Fund.
Kaminsky, Graciela and Sergio Schmukler. 2002. “Short-run Pain, Long-run Gain: The Effects of Financial Liberalization” Unpublished manuscript. World Bank.
Kaplan, Ethan and Dani Rodrik. 2002. “Did Malaysian Capital Controls Work?” in S. Edwards and J. Frankel, eds., Preventing Currency Crises in Emerging Markets. Cambridge, MA: National Bureau of Economic Research.
Karacadağ, Cem, V. Sundararajan and Jennifer Elliot. 2003. “Managing Risks in Financial Market Development: The Role of Sequencing” IMF Working Paper 98/168. Washington, DC: International Monetary Fund.
117
Kim, Soyoung, Sunghyun H. Kim, Yunjong Wang. 2001. “Capital Account Liberalization and Macroeconomic Performance: The Case of Korea” Unpublished manuscript. Available: http://www.tufts.edu/~skim20/paper/kimkimwang.pdf
King, Robert and Ross Levine (1993) “Finance, entrepreneurship and growth,” Journal of Monetary Economics. 32(3): 513-542.
Laurent, Pierre, Nicolas Meunier, Vironique Seltz and Rim Ennajar. 2002. "Capital Outflows: The Flip Side of The Balance of Payments" CDC-IXIS Emerging Regions Report 13. Paris: CDC-IXIS.
Lane, Philip and Gian Maria Milesi-Ferretti (2001), “The External Wealth of Nations: Measures of Foreign Assets and Liabilities for Industrial and Developing Countries,”Journal of International Economics 55: 263–294.
Lucas, Robert E. 1990. “Why Doesn’t Capital Flows from Rich to Poor Countries?” American Economics Review (Papers and Proceedings), 80(2): 92 – 96, May.
McKinnon, Ronald. 1973. Money and Capital in Economic Development. Baltimore, MD: Johns Hopkins University Press.
Mody, Ashoka and Abdul Abiad. 2005. “ HFinancial Reform: What Shakes It? What Shapes It? H," American Economic Review, 95(1): 66-88, March.
Noland, Marcus. 2005. “South Korea’s Experience with International Capital Flows.” NBER Working Paper 11381. Cambridge, MA: National Bureau of Economic Research.
Öniş, Ziya and Ahmet Faruk Aysan. 2000. "Neoliberal Globalization, the Nation-State and Financial Crisis in Semi-Periphery: A Comparative Analysis," Third World Quarterly. 29: 1, 119-139.
Özatay, Fatih. 1999. The 1994 Currency Crisis in Turkey. Unpublished manuscript. Ankara University.
Prasad, Eswar, Kenneth Rogoff, Shang-Jin Wei and M. Ayhan Kose. 2003. Effects of Financial Globalization on Developing Countries: Some Empirical Evidence. IMF Occasional Paper 220. Washington, DC: International Monetary Fund.
Quinn, Dennis P. 1997. “The Correlates of Change in International Financial Regulation,” American Political Science Review, 91(3): 531 – 551, September.
Quinn, Dennis P. And Carla Inclan. 1997. “The Origins of Financial Opennes: A Study of Current and Capital Account Liberalization,” American Journal of Political Science, 41(3): 771 – 813. July.
Rodrik, Dani. 1998. "Who Needs Capital-Account Convertibility?" in Stanley Fischer et al, eds., Should the IMF Pursue Capital-Account Convertibility? Essays in International Finance No. 207, International Finance Section, Department of Economics. Princeton, NJ: Princeton University. May.
Rodrik, Dani and Andres Velasco. 2000. "Short Term Capital Flows" in Annual World Bank Conference on Development Economics 1999. Washington, DC: World Bank.
Rodrik, Dani and Francisco Rodríguez. 2001. “Trade Policy and Economic Growth: A Skeptic's Guide to the Cross-National Evidenceure” in Ben Bernanke and Kenneth S. Rogoff, eds., Macroeconomics Annual 2000. Cambridge, MA: National Bureau of Economic Research.
Rodrik, Dani. 2005. “The Social Cost of Foreign Exchange Reserves” Unpublished manuscript. Harvard University.
Sak, Güven and Fatih Özatay. 2003. “Banking Sector Fragility and Turkey’s 2000-01 Financial Crisis” in Collins, Susan M. and Dani Rodrik, eds., Brookings Trade Forum 2002, Washington, D.C: Brookings Institution Press.
Shaw, Edward. 1973. Financial Deepening in Economic Development. Oxford: Oxford University Press.
Stiglitz, Joseph. 2000. “Capital Market Liberalization, Economic Growth, and Instability,” World Development. 28(6): 1075 – 1086.
UNCTAD. 1999. Trade and Development Report 1999. Geneva: United Nations Conference on Trade and Development.
Williamson, John and Molly Mahar. 1998. A Survey of Financial Liberalization. Princeton Essays in International Finance 211. Princeton University. Princeton, NJ.
Yeldan, Erinç. 2002. “Behind the 2000 / 2001 Turkish Crisis: Stability, Credibility, and Governance, for Whom?” Unpublished manuscript. Bilkent University.
118
APPENDIX
CHRONOLOGY OF LIBERALIZATION EXPERIENCES Argentina Liberalization between 1977 – 82. Average -0,7% growth rate in 1981 – 89
period. “Austral plan” implemented in 1985 temporarily lowers inflation but does little to spur growth. Multiple exchange rates between 1981 – 89. Carlos Menem tooks office in July 1989.
1991: Adoption of “Convertibility Plan”: currency board with US dollar at rate one peso equals one dollar. Complete liberalization of capital accounts. No restrictions on foreign currency deposits in domestic banks. Encouragement of dollarization by making it legal to write contracts in foreign currencies. Reduction in tariffs.
1992: FDI boom to privitized public enterprises. Liberalization and deregulation in many markets.
1994: Bank run after Tequila crisis. Reserve requirements on foreign currency accounts are reduced to inject liquidity to the system.
1997 – 98: Limited contagion from Asian crisis and Russian default.
1999: Brazil devalues real. Loss of competitiveness. Given the real appreciation of peso since the implementation of currency board due to domestic price stickiness, presence of non-tradebles, etc., hit by Brazil’s devaluation have been hard. Fernando de la Rua tooks office in December.
2000: Tax increases to close down the fiscal deficit that stem from excessive expenditure by provinces. Recession instead of a recovery in public accounts. IMF stand-by program.
2001: Exit from currency board in December. All bank accounts are pesofied: 1,4 to 1 in dollar deposity; 1 to 1 in dollar loans, imposing cost on depositors. Bank withdrawals are limited. IMF support withdrawn. Cabinets resigns. Default on debt.
2002 - : effective loophole in limitation of bank withdrawals cause an illusion in balance-of-payments accounts. Argentine residents are allowed to hold more than $1,000 to puchase Argentine stocks. If these Argentine stocks are cross-listed in US markets, they can effectively be sold there and dollar proceeds can be deposited in a US account. Under normal circumstances, these dollar proceed should appear as a capital inflows, as US residents acquire shares in Argentine firms. However, now they appear as outflows. Sources: Ariyoshi et al. (2000), Dominguez and Tesar (2005), Kaminsky and Schmukler (2002)
119
Mexico 1989: Substantial ease of restrictions on portfolio flows. Easing of restrictions on FDI, acquiring of voting shares by non-residents.
1990: Non-residents allowed to hold Mexican government bonds.
1993: Restrictions on borrowing from abroad abolished. Free non-resident access to stock market. Substantial liberalization of FDI.
1994: NAFTA becomes effective. Membership to OECD. Restrictions on investment abroad by investment companies were abolished. Tequila crisis (December). Sources: Ishii and Habermeier (2002), Kaminsky and Schmukler (2002), Williamson and Mahar (1998)
Chile Debt crisis in 1980. Severe restrictions until 1985 and reforms thenafter.
1991: Implementation of encaje with a reserve requirement of 20%. Substantial lessening of restrictions on investment by residents abroad.
1992: Extention of encaje base and increase of the rate to 30%. Residents are allowed to issue bonds abroad. First allowence for pension funds to invest abroad, albeit with low ceilings.
1994: Ease on restrictions of pension funds investment to abroad.
1997: Further lessening of restrictions on resident’s investment abroad.
1998: Encaje are eliminated by setting the reserve requirement rate at 0%. Funding obtained from bonds issued abroad by residents are no longer have to returned to Chile.
1999: Substantial lessening of restriction on investment of pension funds abroad. Sources: Ishii and Habermeier (2002), Cowan and De Gregorio (2005), Johnston et al. (1997), Kaminsky and Schmukler (2002).
Brazil A modest opening of capital account with sporadic applications of capital controls. Debt crisis in early 1980s. Interruption of capital flows after 1983.
1988: Official regulation of parallel exchange rate markets, causing an incentive to broaden transactions through parallel markets.
1990: Certain financial institutions were authorized to obtain resources from abroad through the issuance of commercial papers.
1991: Several substantial measures to liberalize capital flows.
1993: Implementation of quantity and price-based restrictions on inflows.
1994: Real plan to reduce inflation.
1995: Temporary ease of controls to mitigate the effects of Tequila crisis.
120
1996: Tightening of controls. Constant evolution of controls as market participants find ways to circumvent them.
1997: Controls on inflows relaxed as a response to Aisan crisis.
1999: Devaluation of real. Exchange rates allowed to float. Elimination of multiple exchnage rates. IMF-supported program.
2000: Non-residents are allowed to invest under the same rules with residents. Sources: Goldfajn and Minella (2005)
Peru 1970 – 85: Recurring balance of payments crises, low growth rates, high inflation, fiscal imbalances.
1980: Dual exchange rate regime.
1985: Stabilization program reduced fiscal deficit, however slow growth rates remain. Extensive controls on capital flows. High inflation in 1989.
1990: Reform program including liberalization of current account, capital account; exchange rates unified and allowed to float, supervisory framework enforced on banks,
1994: Elimination of differential treatment of resident and non-resident agents by constitutional amendment. Sources: Ariyoshi et al. (2000)
Columbia 1991: “Apertura program” unification fo exchange rates, controls on borrowing abroad relaxed. Some controls on capital inflows remain, however.
1992: A 10% witholding tax aiming to reduce certain speculateive cross-border transactions. Residents were allowed to hold portfolio investments abroad up to 500,000$.
1993: Implementation of Chilean-style URRs at a reate of 47%. Continued surge in inflows. Effects of URRs are mixed. They may have played a role on lengthened maturity of debt and conversion of some short-term flows to FDI.
1994: URR rates were raised to a range of 43% to 140%.
1997 – 98: Lowering of the rate of URRs.
2000: Elimination of URRs. Sources: Ariyoshi et al. (2000), Ishii and Habermeier (2002)
Venezuella 1989: Virtually all forms of exchange rate controls abolished, free capital accounts regime.
1994 – 96: Extensive controls on both inflows and outflows. Multiple
121
exchange rate markets, with parallel market premium of 40 – 100%.
1996: Implementation of IMF program and liberalization back.
2003: Imposition of controls on both current and capital account transactions. Source: IMF (2005), Ariyoshi et al. (2000), Kaminsky and Schmukler (2002)
Korea Financial repression before 1980.
1982: Restrictions on foreign borrowing under $1 million eased.
1985 – 1990: Several minor measures of liberalization.
1991: Residents were allowed to issue securities in foreign currencies to finance imports of inputs and machinery for which no domestic substitute available. Nonresidents who acquired Korean shares through convertible bonds were allowed to trade them in the stock exchange. Nonresidents were allowed to convert up to $100,000 to invest with a maturity of more than 2 years.
1992: Residents were allowed to issue abroad negıtiable certificates of deposit and commercial paper, and maintain accounts abroad. The stock exchange was opened to nonresidents with quantitative limitations.
1993: Announcement of capital account liberalization plan. Non-residents were allowed to held local currency accounts. The list of institutional investors who can invest abroad was expanded to include investment companies and pension funds.
1994: Ceilings on borrowing from foreign financial instititutions abolished. Direct overseas stock investment allowed witha ceiling.
1995: Relaxation of foreign ownership restrictions on Korean firms from 10% to 12%
1996: Membership to OECD. Nonresidents allowed to hold local currency accounts in domestic banks.
1997: Asian crisis Sources: Ishii and Habermeier (2002), Johnston et al. (1997), Noland (2005), Kaminsky and Schmukler (2002).
122
Malaysia Capital account mostly liberalized in 1970s. Furher deregulation of FDI and portfolio inflows in mid-1980s.
1994: Imposition of temporary controls on inflows, including a ban on selling of Malaysian securities of less than one-year maturity to non-residents by residents, ban on commercial banks to engage in non-trade-related bid-side swaps and forward transactions with non-residents and asymmetric open position limits on banks.
1995: Relaxation of controls on inflows.
1997: Asian crisis.
1998: Imposition of temporary controls on capital outflows and peg of local currency to dollar. Outflows of portfolio investment were obliged to be held fro a one-year period.
1999: Ease of controls on outflows by replacing one-year period with a gradual scheme, allowing profits to be repatriated by paying an exit tax.
2001: Controls on outflows are completely abolished. Sources: Williamson and Mahar (1998), Ariyoshi et al. (2000), IMF (2005), Kaplan and Rodrik (2002)
Thailand Relatively liberal capital account transactions even before the major reforms are implemented.
1985 – 87: Liberalization of capital inflows in the form of portfolio investment, restririctions on portfolio outflows remain, however.
1991: Allowence to lend by residents to companies abroad which has at least 25% Thai participation. Puchases of assets more than 10$ million are subject to approval.
1993: Opening of Bangkok International Banking Facility, which facilitated borrowing from abroad to a large extent.
1995: Response of monetary authorities to surge in capital inflows by imposing 7% reserve requirement at the Central Bank.
1997: Asian crisis begins at Thailand. Imposition of controls on outflows, by restricting lending in local currency to non-residents.
1998: Controls on outflows are lifted. Sources: Ariyoshi et al. (2000), Ishii and Habermeier (2002), Kaminsky and Schmukler (2002), Johnston et al. (1997).
Indonesia 1985: Liberalization of outflows by residents individuals; controls on outflows by banks and financial institutions remain.
1989: Complete liberalization of portfolio capital inflows.
123
1991 – 96: Several quantitative controls on offshore borrowing and government borrowing from abroad.
1997: Asian crisis. Source: Johnston et al. (1997)
Philippines Limited liberalization in 1980s. Gradual elimination of all restrictions on capital account between 1992 – 95. Source: Williamson and Mahar (1998)
Hungary 1989: Liberalization of FDI by non-residents, along with current account.
1993 – 95. Minor liberalization attempts on capital flows.
1996: OECD membership. However trasitional clauses are applied for some OECD codes on capital flows. Commercial credit transactions with non-residents, opening foreign exchange accounts at domestic commercial banks by residents and personal capital movements are allowed.
1997: Purchase of OECD member government bonds by residents allowed.
1998: Russian crisis. Short-lived effect on Hungary.
2001: Full application of OECD codes on capital flows, complete liberalization. Source: Ishii and Habermeier (2002)
Czech Rep.
1993 - 1995: Rapid liberalization of capital flows. Some restrictions remain on mostly outflows.
1995: OECD membership. Implementation of a foreign exchange transaction fee to limit inflows, which remain ineffective.
1997: Slow down in capital inflows, currency crisis and implementation of a macroeconomic policy package for recovery.
1999: Lifting of limits on foreign exchange transactions.
2001: Issuence of debt instruments abroad by residents allowed.
2004: Membership to EU and full liberalization. Source: IMF (2005)
Poland 1993 – 95. Minor liberalization attempts on capital flows.
1995: Application of IMF Article VIII. As a result, most restrictions on capital flows are removed.
1996: OECD membership
2004: EU membership Source: de Souza (2004)
124
125